The Pennsylvania State University The Graduate School College of the Liberal Arts
AN EXAMINATION OF LEADERSHIP BELIEFS AND LEADERSHIP SELF-IDENTITY: CONSTRUCTS, CORRELATES, AND OUTCOMES
A Thesis in Psychology by Nathan J. Hiller © 2005 Nathan Jeffrey Hiller Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy December 2005
UMI Number: 3202503
UMI Microform 3202503 Copyright 2006 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, MI 48106-1346
The thesis of Nathan J. Hiller was reviewed and approved* by the following:
David V. Day Professor of Psychology Dissertation Adviser Chair of Committee
James L. Farr Professor of Psychology
Donald C. Hambrick Smeal Chaired Professor of Management
Susan Mohammed Associate Professor of Psychology
Kevin R. Murphy Professor of Psychology Head of the Department of Psychology
* Signatures are on file in the Graduate School.
iii ABSTRACT The way we think about ourselves as leaders and what we believe leadership to be are important guides of subsequent thoughts and actions in the leadership domain. In two separate studies, measures of self-representations in the domain of leadership (leadership self-identity) and views about the nature of leadership (orientation toward leadership) were developed and examined in a sample of undergraduates and a sample of medical center employees. Among the student sample, possessing a leadership self-identity was related to previous leadership experience, core self-evaluations, motivation to lead, and selfmonitoring. The three dimensions of leadership orientation (dominance, development, and shared) were differentially related to individualism and collectivism dimensions. In the medical center sample, results of hierarchical linear modeling showed no main effects of supervisor leadership self-identity on leadership potential or leader-member exchange. In several cases, however, leadership self-identity of supervisors interacted with leadership orientation in predicting ratings of subordinate leadership potential. Interest in leadership development was related to leadership identity and leadership orientation through both main and interaction effects. These results provide some evidence that both self-identity around leadership and cognitive beliefs about the nature of leadership need to be examined in order to better understand leadership phenomena.
iv TABLE OF CONTENTS
List of Tables. .…………………………………………………………………………….. vii List of Figures…………………………………………………………………...……......... ix Chapter 1
Introduction: Seeing the Self as a Leader - An Examination of Leadership Self-Perceptions …………………...……………………………………........
1
The Importance of Leadership……………………………………………………….. The Importance of Self Views……………………………………………………….. General Research on Self-Schemas………………………………………………….. Stable and Temporally Influenced Components of the Self…………………………. Self-Schema Measurement…………………………………………………………... What is Known About Leadership Self-Schemas………………………………….... Leadership Self-Identity……………………………………………………………...
1 2 3 4 4 7 9
Chapter 2 Origins and/or Correlates of Leadership Self-Identity………………...………
10
Experience…..……………………………………………………………………......... Personality Correlates of Leadership Self-Identity……………………………………. Core Self-Evaluations……………….……………………………………………….. Motivation to Lead………………………………….……………………………….. Self-Monitoring……………………………………….…………………………....... Pattern Matching: Matching Self-Traits to Traits of Leader……………………………
10 11 11 12 13 15
Chapter 3 Individual Theories about the Nature of Leadership…………………………..
16
Orientation Toward Leadership……………………………………………………....... Level of Self-Construal (Individualism and Collectivism)…..………………….…….. Individualism Dimensions…………………………...…………………………....... Collectivism Dimensions……………….……………………...……………………
16 19 20 21
Chapter 4 Outcomes of Leadership Self-Identity and Orientation Toward Leadership....... Leader-Member Relationships………………………….………………………............. Leadership Potential……………………………………...……………………………. Orientation Toward Leadership as a Moderator………………………………......... Behavioral Repertoire…………………………………………………………….......... Interest in Leadership Development ………………………………………………....... Does Leadership Self-Identity Always Matter? – Boundary Conditions…………........ Two Proposed Studies…....…………………………………………………………….
23 23 23 24 24 26 28 28
Chapter 5 Development of the Leadership Self-Identity and Orientation Toward Leadership Measures……..……………………………………………............
30
Development of the Leadership Self-Identity Scale………………………..………….. Development of the Orientation Toward Leadership Measure………………….…….. Measure Refinement………………………………………………………….…….......
30 31 36
v Refinement of the Leadership Self-Identity Measure…………………...………….. 36 Refinement of the Orientation Toward Leadership Measure……………………...... 38 Chapter 6 Study 1: Method, Results, and Discussion……………………………………. Method…….……………………………………………………….………………....... Participants.……………………………………………………….………………… Procedure.……………………………………………………….………………….. Measures.……………………………………………………….…………………... Results…..……………………………………………………...……………………… Testing Antecedents/Correlates of Leadership Self-Identity in a Single Model...…. Orientation Toward Leadership…………………………………………………….. Discussion…..………………………………………………………………………….. Leadership Self-Identity…………………………………………………………….. Orientation Toward Leadership………………...………………………………....... Relationship between Leadership Self-Identity and Orientation Toward Leadership Brief Summary…………………………………………………………………………
50 50 50 50 51 54 58 59 59 59 62 63 64
Chapter 7 Study 2: Method Results, and Discussion.….………………………………… Method…………………………………………………………………………………. Participants..………………………………………………………………………… Procedure..………………………………………………………………………….. Measures..………………...………………………………………………………… Results…..……………………………………………………………………………... Testing Hypotheses 8, 9, and 10 Using Hierarchical (Multi-level) Models……...… Hypotheses 11-14…………………………………………………………………… Discussion……………………………………………………………………………… Overview……………………………………………………………………………. Differences Between Subordinates and Supervisors……………………………....... Effects of Supervisor Leadership Self-Identity and Orientation Toward Leadership on LMX and Leadership Potential Ratings……….…………...…………...…..... Interest in Leadership Development………………………………………………...
65 65 65 66 68 69 74 86 92 92 92 93 95
Chapter 8 General Discussion…………………………...…………………………..........
97
Leadership Self-Identity and Orientation Toward Leadership Measures……………… Relationship Between Leadership Self-Identity and Orientation Toward Leadership……………………………………………………………………............ Limitations and Future Research…..………………………………………………....... Implications……………………………………………………………………………. Conclusion………………………………………………………………………….......
99 100 100 103 104
References………………………………………………………………………………….. 105 Appendix A Orientation Toward Leadership: The 50 Initial Items Developed.………….. 118
vi Appendix B Q-Sort Task: Instructions Given to Raters, Description of Dimensions, and Items Assigned to Their Intended Construct……………………………………. 121 Appendix C Item-Level Descriptive Statistics for Study 1.………………………………. 124 Appendix D Item-Level Descriptive Statistics for Study 2..…………………………........ 131 Appendix E Survey used for Student Sample.………………………………………......... 135 Appendix F Factor Analysis Results of Leadership Experience Measure.……………….. 152 Appendix G Pre-Survey Letter Sent to PSHMC Participants.……………………………. 155 Appendix H Survey Used for PSHMC Supervisor Sample………………………………. 157 Appendix I Survey Used for PSHMC Subordinate Sample…………………………....... 162 Appendix J Summary Explanation of Research for PSHMC Participants.……………… 167 Appendix K Follow up Letter Sent to PSHMC Participants……………….…………....... 170
vii LIST OF TABLES Table 1 Results of Item-to-Construct Assignment Task…………………………………... 34 Table 2 Exploratory Factor Analysis of Leadership Self-Identity Items in Studies 1 and 2. 37 Table 3 Results from Principal Axis Factoring of Orientation Toward Leadership Items: Study 1……………………………………………………………………………………. 40 Table 4 Results from Principal Axis Factoring of Orientation Toward Leadership Items: Study 2……………………………………………………………………………………. 43 Table 5 Correlations Between Resultant Orientation Toward Leadership Factors: Study 1 and Study 2………………………………………………………………………. 45 Table 6 Confirmatory Factor Loadings for Leadership World-View Items……………..... 47 Table 7 Fit Statistics for Confirmatory Factor Analyses Comparing the Fit of Single-Factor Versus Two Alternative Three-Factor Models…………………………….. 48 Table 8 Correlations Among Variables for Student Sample……………………………..... 56 Table 9 Multiple Regression Results of Previous Leadership Experience Dimensions Predicting Leadership Self-Identity……………………………………….... 57 Table 10 Multiple Regression Results Examining Antecedents to Leadership SelfIdentity: Descriptiveness, Certainty, and Importance…………………………………….. 58 Table 11 Supervisor and Subordinate Means on Leadership Self-Identity, Orientation Toward Leadership and Leadership Development Items…………………………………. 70 Table 12 Correlations Among Common-Measured Variables for Medical Center Sample (Supervisors and Subordinates)…………………………………………………... 71 Table 13 Correlations Among All Supervisor Assessed Variables: Medical Center Sample…………………………………………………………………………………….. 72 Table 14 Correlations Among All Subordinate Assessed Variables: Medical Center Sample…………………………………………………………………………………….. 73 Table 15 Hierarchical Linear Model (HLM) Results Predicting Supervisor Ratings of Subordinate-Level LMX with Leadership Self-Identity Descriptiveness………………… 77 Table 16 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate-Level LMX with Leadership Identity Certainty…………………………….. 78
viii Table 17 Hierarchical linear model (HLM) Results: Predicting Supervisor Ratings of Subordinate-Level LMX with Leadership Identity Importance…………………………... 78 Table 18 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate-Level LMX with All Three Leadership Identity Dimensions……………..... 79 Table 19 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate Leadership Potential with Leadership Identity Descriptiveness…………….. 80 Table 20 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate Leadership Potential with Leadership Identity Certainty………………….... 80 Table 21 Hierarchical Linear Model (HLM) Results: Predicting Supervisor-Ratings of Subordinate Leadership Potential with Leadership Identity Importance………………..... 81 Table 22 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate Leadership Potential with All Three Leadership Identity Dimensions…….... 82 Table 23 Hierarchical Linear Model (HLM) Results: Predicting Leader-Member Exchange Quality with Development Orientation Toward Leadership…………………... 83 Table 24 Supervisor Ratings of Subordinate Leadership Potential with LSI Descriptiveness and Dominance Orientation……………………………………………... 84 Table 25 Supervisor Ratings of Subordinate Leadership Potential with LSI Importance and Dominance Orientation………………………………………………….. 85 Table 26 Correlations Between Leadership Self-Identity of Supervisor and Subordinate Ratings of Supervisor Behavioral Repertoire……………………………………………... 87 Table 27 Summary of Significant Moderated Regressions: Predicting Interest in Leadership Development with Leadership Self-Identity and Orientation Toward Leadership………………………………………………………………………………… 89
ix LIST OF FIGURES Figure 1 The three orientations toward, or world views of, leadership (based on Drath, 2001)…………………………………………………………………… 18 Figure 2 Quinn’s competing values framework..….………………………...……..……. 26 Figure 3 Overall model of proposed antecedents and outcomes of leadership self-identity and orientation toward leadership: Study 1 and 2.………….………………. 29 Figure 4 Measure of Leadership Self-Identity…………………….….…….……….......... 31 Figure 5 Scree plot of eigenvalues by Orientation Toward Leadership factor: Study 1….. 39 Figure 6 Scree plot of eigenvalues by Orientation Toward Leadership factor: Study 2 ……………………………………………………….…………………... 42 Figure 7 LISREL 8 correlations between factors for four-correlated factors model……. 48 Figure 8 Example of nested nature of supervisor-ratings of subordinate leadership potential………………………………………………………………………
75
Figure 9 Subordinate leadership potential by supervisor Leadership Identity Descriptiveness.…………………………………………………………………………... 84 Figure 10 Subordinate leadership potential by supervisor Leadership Identity Importance………………………………………………………………………………... 85 Figure 11 Interest in leadership development by Leadership Identity Importance: Supervisors and subordinates......…………………………………………… 90 Figure 12 Interest in leadership development by Leadership Identity Certainty: Supervisors and subordinates………...…………………...………………....................... 91
1 CHAPTER 1 Introduction: Seeing the Self as a Leader An Examination of Leadership Self-Perceptions A significant amount of work in the leadership domain has taken a social-cognitive approach to examining perceptions in the leadership process, though most of this work has focused on follower perceptions of leaders, leader perceptions of followers (see Lord & Maher, 1991, for a primer), and more recently, leader effects on follower self-identity (Lord & Brown, 2004, Lord, Brown, & Freiberg, 1999). Yet for all of the attention to socialcognitive processes, almost no attention has been devoted to how individuals in organizations perceive themselves as leaders, and what the effects of leadership self-perceptions might be. This dissertation takes the perspective that understanding the “self” is an important component to understanding leadership. The self, which acts as the meeting point of cognitions, personality, and social factors, provides a general basis for interpreting the world and producing behavior (e.g., Cross & Markus, 1994; Gabriel & Gardner, 1999; Yurek, Farrar, & Andersen, 2000) and thus is a promising avenue for understanding the leadership process in a wide variety of situations (Lord et al., 1999). This dissertation research seeks to better understand views of the self as leader, relate self-perceptions of leadership identity to other important variables (theoretically and empirically), and explore the outcomes of leadership self-identity. Central questions to be examined include: what kinds of people are more likely to view themselves as leaders and how does leadership self-perception influence organizationally relevant behaviors? In addition to better understanding how people think of themselves as leaders (or not), another major goal of this research is to understand different orientations toward leadership; what people believe leadership is. Leadership research, up to this point, has sought to understand these leadership “theories” by studying individual perceptions about traits and qualities that are associated with being a leader (see Schyns & Meindl, 2005, for a recent overview). I assert that understanding leadership “theories” requires more than simply an understanding of traits that people believe are associated with leaders. Leadership involves actions and processes, and individuals are likely to have differing views about what actions and processes are important for leadership. These fundamental orientations toward leadership are in turn likely to guide leadership behaviors and perceptions of others’ leadership potential, and may interact with leadership self-identity in leading to important outcomes. Central questions to be examined include: how do different individuals think about the nature of leadership, and how do one’s perceptions about the nature of leadership interact with leadership self-identity? The Importance of Leadership Whether stated explicitly or taken as an implicit assumption, almost all organizations desire an abundant resource of leaders and leadership. Leadership, as the “influential
2 increment over and above mechanical compliance with the routine directives of the organization” (Katz & Kahn, 1978, p. 528), provides a stage from which organizations can adapt, survive, and thrive in an increasingly competitive and changing world (D’Aveni, 1994; Raelin, 2004). It is not at all surprising then, that the study of leadership is of central importance to those who desire to understand how organizations of all kinds might function more effectively. In the general literature on leadership, much of the recent research has been focused on the effects that leaders have, both in terms of subordinate behaviors1, and group and organizational outcomes (e.g., Dvir, Eden, Avolio, & Shamir, 2002; Peterson, Smith, Martorana, & Owens, 2003). Other research has considered the differences between leaders and non-leaders, or effective versus ineffective leaders. This general stream of research has examined characteristics of leaders such as personality traits (e.g., Judge, Bono, Ilies, & Gerhardt, 2002; Lord, De Vader, & Alliger, 1986; Lowe, Kroeck, & Sivasubramaniam, 1996; Stogdill, 1974) or behaviors of leaders (e.g., Bass, 1985; Yukl & Lepsinger, 1990; Ohio State and Michigan leadership studies). Each of these streams of leadership research, in placing the emphasis of study in slightly different places, has contributed greatly to our understanding of leadership. Yet while increasingly more is known about the behaviors and personality characteristics (as well as outcomes of these behaviors and traits) that are associated with leadership, very little is known about the mind of a leader, how leaders view themselves, and how their self-views (and theories about leadership) are structured. Consistent with research in social, cognitive, and clinical psychology suggesting the importance of self-perceptions for both thought and action (discussed in more detail below), leadership self-identity (or self-schema) is viewed as an important part of understanding leaders and leader behaviors. The Importance of Self Views One of the most basic and important findings in cognitive psychology is the overwhelming evidence that humans possess and constantly use interpretive frameworks in order to understand the world. Broadly known as schemas, these structures provide us with the frameworks of how the world operates and are necessary for almost everything that we do. We have an almost immeasurable number of schemas, which enable us to do things like identify and categorize objects or make quick adjustments while swinging a baseball bat. In social situations, these interpretive frameworks are equally essential; they help us, for example, carry on conversations with others (through schemas about how conversations should progress, how sentences are structured, and the appropriateness or inappropriateness of certain topics). In our perceptions about others, we also have schemas. One of the ways we describe and make sense of others is by placing them into social categories (e.g., bluecollar laborer, housewife, leader) or labeling them as having certain traits (e.g., spontaneous, extraverted, intelligent, aggressive). But just as traits and social categories are used to understand others, they are also applied to thoughts and descriptions about oneself (Lord & 1
The word “subordinates” is used here for maximum clarity in differentiating between supervisors and supervisees. This word, as used in this dissertation research, does not indicate lower worth or importance.
3 Brown, 2004). Over a period of time, as individuals come to “know” themselves through observations of thoughts and behaviors, as well as reflections of the self from others2, “selfschemas” are developed that organize memory and behavior (Kihlstrom & Klein, 1994). These self-schemas are particularly important because, once formed, they have a further influence on our subsequent thoughts and behaviors (Cross & Markus, 1994; Markus, 1977). General Research on Self-Schemas Originally defined by Markus (1977), self-schemas are cognitive generalizations about the self that organize and guide the processing of self-related information.3 In essence, self-schemas represent the existence and salience of self knowledge in a given domain. Schematics (the term given to an individual with a self-schema in a given domain) are aware of themselves in this domain and assign this self-knowledge critical personal importance, whereas aschematics are neither particularly aware of themselves nor do they assign this selfknowledge critical importance. A self-schema may include representations in virtually any domain such as “I am cautious,” or “I am usually outgoing in new social situations,” or “I am a leader.” Because a self-schema in a given domain plays an important part of that individual’s identity, schematic individuals pay close attention to, and favor information in their schematic domain (Markus, Cross, & Wurf, 1990). Not surprisingly, schematic individuals also interpret events according to the lens of their schema, believe they are capable in that domain, and will seek out opportunities that allow them to demonstrate their self-view. The domain specificity of self-schemas means that different self-schemas are relevant to corresponding cognitions and behavior across an inumerable number of domains. It is not surprising, then, that dozens of empirical investigations have focused on self-schemas in areas ranging from problem solving and academic achievement (e.g., Cross & Markus, 1994; Tarquinio & Somat, 2001), depression (e.g., Dozois & Dobson, 2003; Penland, Masten, Zelhart, Fournet, & Callahan, 2000), sexuality (e.g., Andersen, Cyranowski, & Espindle, 1999; Cyranowski & Andersen, 2000), fitness (e.g., Estabrooks & Courneya, 1997; Kendzierski, Sheffield, & Morganstein, 2002), religiosity (Morrison & Morrison, 2002), pain management (e.g., Pincus & Morley, 2001), prosocial behavior (e.g., Froming, Nasby, & McManus, 1998), and responses to advertising (Forehand, Deshpandé, & Reed, 2002). In these different empirical investigations, it has been found that when a well-articulated generalization about the self is present (i.e., an individual has a self-schema in a given domain), a consistent pattern can be found in that domain with respect to one’s decisions and behaviors, as well as one’s assessments of others. In addition, self-schematics in a given domain are more resistant to change than aschematics (Markus, 1977; Petersen, Stahlberg & Dauenheimer, 2000) and schematic individuals are more likely to make external attributions 2
There are philosophical problems in presuming that there is a unique ‘self’ that is somehow able to reflect on the same self as an abstract observer, however for practical purposes (and consistent with most of psychology), the conceptual difficulties inherent in ‘dualistic’ thinking are glossed over. For a good introduction to this topic, see Slife and Williams (1995). 3 Self-schemas are sometimes also referred to as salient identities or core self-structures (Gergen; 1968; Stryker, 1980), though self-schema is the predominant term and thus will be used in this work.
4 for their own behavior when their behavior doesn’t match their schema (e.g., Kendzierski et al., 2002). Self-schemas provide a lens through which information (behavioral and nonbehavioral) is perceived and act as a guide for future behaviors, perceptions and attitudes. Cross and Markus (1994) suggest that without well elaborated self schemas, individuals may experience difficulty maintaining a high level of performance on a task over time, even if they have the required competencies. Unless aschematic persons begin to construct a self-schema in this domain with which their experiences and abilities can be organized and directed, they will be less likely to attend to their abilities and will choose more personally satisfying activities. Stable and Temporally Influenced Components of the Self This dissertation aims to understand consistent patterns of the self (self-schemas) and concomitant behaviors across both situations and time. At the same time, however, it is acknowledged that in a given situation, certain portions of the self are likely to be more highly activated than others. In order to understand behaviors and information processing at a given point in time, it is first important to understand that the self is multidimensional. One individual may be a prison guard, volleyball player, romance-novel enthusiast, parent, and death-penalty activist. This person may also be loving, fair, warm, and logical. Different portions of the selfconcept are activated at different times, and these activated self-concepts guide action and information processing on a moment-by-moment basis (Lord, Brown, & Freiberg, 1999; Kihlstrom & Klein, 1994). First explicated by Markus and Wurf (1987) this working selfconcept (WSC) is a continually shifting conglomeration of core self-views and more peripheral aspects of the self made salient (i.e., activated) by context. However, while there appears to be evidence for a moment by moment nature of the WSC, the self does not just blow with the wind (Judge, Erez, Bono, & Thoresen, 2003); certain components of the self are much more likely to be activated across a variety of situations (Lord & Brown, 2004). While the WSC that is primed at any given moment is an excellent proximal predictor of specific behavior (just as specific attitudes are the best predictors of specific behaviors), stable differences in the self-concept are better predictors of relatively broad behavioral patterns and behavioral patterns across situations. Given the interest in predicting leadership across situations, the focus of this dissertation is on the stable components of the leadership self-concept, rather than the temporally primed component of the self-concept (WSC). Self-Schema Measurement There have been three basic methods for assessing self-schemata, two of which involve pencil and paper self-assessments (direct and indirect) and one measuring response latency using a computer. Within the leadership domain, only the indirect pencil and paper
5 method has been used to assess self-schemas. The commonality among the different approaches is that participants are asked to assess descriptors or items in relation to their self. Yet while related to each other, correlations between different methods of assessment suggest that they are not equivocal and thus an appropriate understanding of leadership self-schemata should carefully consider theoretical reasons for method of assessment. Each method, as well as a more detailed examination of the relationship between methods, will be described below. Two Pencil and Paper Methods – Direct and Indirect. Before discussing discrepancies between the two pencil and paper methods of assessment, a general discussion of commonalities between methods is important. Two common advantages of the direct and indirect pencil and paper method are that it is relatively simple to administer and relatively easy for participants to complete. Both methods require participants to assess items (on a Likert-type scale) first in terms of their self-descriptiveness, and then in terms of the importance of each statement to their overall self-evaluation. In the classic definition (Markus, 1977), for an individual to be considered “schematic” in a certain domain, he or she must assess a statement as self-descriptive and assign that trait or ability high personal importance. Individuals who are not assigned to the “schematic” category, are not, however, automatically considered aschematic. Aschematics assign moderate ratings of selfdescriptiveness and low to moderate assessments of self-importance to items (e.g., Catrambone & Markus, 1987; Cross & Markus, 1994; Markus, 1977). Once schematics are distinguished from aschematics, between-group comparisons are made. Far from being operationalized in only one way, recent direct and indirect empirical pencil and paper assessments of self-schematicity (including papers by Markus) have used variants on the classical measurement and definition of schematic and aschematic individuals (e.g., Avants, Margolin, & Kosten, 1996; Forehand, Deshpandé, & Reed, 2002; Froming, Nasby, & McManus, 1998). In the more general literature on self-views that has evolved as a result of Markus’ original work, researchers have often considered the components of selfviews (self-descriptiveness and importance of this self-view) separately and in interaction with each other. Further, it has also been shown that certainty of self-description is a unique and important part of the self-concept in a given domain (e.g., Pelham & Swann, 1989; Swann & Ely, 1984; Swann & Pelham, 2002). Thus, a self-schematic (generally) would have a high self-awareness, be certain of his/her self-assessment, and assign his/her self-view critical self-importance. But, perhaps more importantly, these different aspects of self-views in a given domain (descriptiveness, importance, and certainty) are conceptually distinct and need to be examined separately and in interaction. In other words, a simple distinction between schematics and aschematics may obscure important and interesting nuances. For simplicity, the current research will generally frame hypotheses simply, in terms of selfschematics vs. non-schematics, with the realization that analysis of nuances of these different components (descriptiveness, certainty, and importance) may also be informative. The direct and indirect pencil and paper methods are highly similar and are not considered separately in the literature on self-schema. These similarities, however, should not obscure an important difference in the conceptual clarity of the measures. The direct
6 method of assessment requires participants to first assess the self-descriptiveness of a series of related statements on a Likert-type scale, followed by an assessment of the certainty of that view and the self-importance to one’s identity. Items assessing a child’s mathematical self-schema, for example, might include statements such as “I am intelligent” and “I am good at math,” and respondents would assess the descriptiveness, certainty, and importance of each item. Compared to the indirect approach (described below), there is less guesswork and a smaller “inferential leap” as to what is actually being assessed in the measure. The indirect pencil and paper method approaches the domain of interest from a slightly more veiled way in that this method does not directly inquire about the domain of interest. In the literature on self-schemas, a salient example occurs in the single study on leadership self-schemas. Engle and Lord (1997) used the indirect pencil and paper approach in an assessment of leadership self-schema by assessing the self-descriptiveness and self-importance of a variety of traits associated with leadership (e.g., dominance, intelligence, masculinity) without directly asking about their self-representation in the leadership domain. The traits used to assess leadership self-schematicity were those traits that were found to be generally held implicit leadership traits. More specifically, traits that received a high endorsement from the entire sample as being general “leadership” traits were then used to assess personal leadership self-schemas by asking participants to rate themselves on those same traits. Indirect methods of self-schema assessment have been used by other researchers as well, particularly when the self-schema of interest is in a complex or highly sensitive domain (e.g., communion and agency domains of interpersonal belief; Froming, Nasby, & McManus, 1996; sexual self-schema; Reissing, Binik, Khalifé, Cohen, & Amsel, 2003). This method could be less prone to social desirability effects than direct assessment, and may be particularly useful when inquiring about highly loaded or difficult-to-breach topic such as prejudice or sexual orientation. Latency Response Method. The third approach to assessment of self-schemas is a latency response technique and is probably better thought of as a measure of working selfconcept (WSC). The basic rationale for this approach is that the self, as a highly organized cognitive structure (Aron, Aron, Tudor, & Nelson, 1991) is able to more quickly recognize traits that are either highly self-descriptive or highly non-descriptive than traits that are of neutral descriptiveness. Research using this technique typically requires participants to look at a computer screen while a series of traits or descriptors are presented (one at a time). Participants are asked to judge, as quickly as possible, whether or not the traits or descriptors are characteristic of themselves. Because domains for which an individual does not hold a readily accessible schema are more difficult to access, longer response latencies are thus indicative of a weaker schema. This interpretation, however, is potentially problematic because of the existence of a counteracting effect, known as the distinctiveness effect. The distinctiveness effect occurs when traits distinctive to the self are recognized more slowly than traits that describe many
7 other people (Mueller, Ross, & Heesacker, 1984). Thus, a slow response may be indicative of not holding a schema, or it may be indicative of a perception that a given trait is not commonly held. Another difficulty in interpreting latency results is that processing speed and reaction time are heavily dependent on the temporary working self-concept, rather than being indicative of more stable, underlying self factors. Indeed, correlations between response latencies and pencil and paper measures of self-schema are generally only moderate (Markus, 1977), and endorsement of survey items (either on a yes/no or a Likert-type scale) have generally been favored in recent investigations of stable individual differences in selfschemas. What is Known About Leadership Self-Schemas In the single published empirical assessment of leadership self-schemas, Engle and Lord (1997) examined leadership self-schemas of both supervisors and subordinates in a marketing department of an electric company. In addition to leadership self-schemas, implicit leadership and performance theories, leader-member exchange quality and affectivity were also measured. Leadership self-schemas were assessed through the indirect pencil and paper method (as mentioned above), and measurement involved a two-stage process. First, implicit leadership theories were assessed by asking participants to rate how prototypical each of 23 traits or behaviors is to leadership. Only items receiving a grand mean of 4 or higher on a 5-point Likert type scale were considered prototypical of leadership and examined further. This resulted in 10 “typical” leadership theory traits that were widely endorsed across the organization: intelligent, cooperative, enthusiastic, decisive, sincere, goal-oriented, persuasive, wise, dedicated, and motivated. After the highly endorsed implicit leadership traits were determined, leadership self-schema was then assessed using only those 10 implicit leadership adjectives. As a result of this method, leadership self-schema in this study was indicative of the extent to which the organizationally defined leadership prototype was descriptive of an individual’s self-concept. Several results from this study of 18 supervisors and 76 subordinates are worth noting. First, both supervisors and subordinates who were self-schematic for organizationally endorsed leadership attributes also had self-schemas for organizationallyendorsed effective performance. In other words, individuals who saw leadership adjectives as more self-descriptive also saw effective performance adjectives as more self-descriptive. A significant positive relationship was also observed between an individual’s leadership selfschema score and the extent to which he or she endorsed organizationally-defined leadership and performance traits and behaviors (implicit leadership and performance theories). Leadership self-schema was also positively related to a measure of positive affectivity and negatively related to negative affectivity. Additionally, and perhaps at first glance somewhat surprisingly, the more a supervisor held an organizationally normative leadership self-schema (i.e., the higher the self-schema score), the less they liked their subordinates and the less they had a positive leader-member exchange (LMX) relationship with their subordinates. One interpretation of this finding is that self-schematic supervisors (who have highly developed cognitive
8 structures around leadership) are more stringent judges. Previous research in the area of selfschemas has shown that being self-schematic in a domain provides a chronic and habitual self-referent (see Cross & Markus, 1994), and when one references the self in making judgments about others, judgments are more stringent (Dunning & Hayes, 1996). However, the position taken in this dissertation is that while having a leadership selfschema was found in the Engle and Lord (1997) study to be related to poorer quality LMX, this may partly be an artifact of the indirect measure of self-schema and the fact that beliefs about leadership were ignored. It is important not only to understand how people see themselves as leaders, but also what they believe leadership is. I argue that being schematic with regard to leadership does not necessarily lead to poorer LMX, an issue that will be discussed in more detail later in this paper. It is even possible that leadership-schematic supervisors may have better relationships with subordinates when their views of leadership include a more collective view of what leadership is. This would mean that the negative relationship between supervisor leadership self-schema and supervisor-rated LMX are moderated by the supervisor’s orientation, or “theory”, about what is included in leadership. A direct measure of leadership self-schema (e.g., “Leadership Self-Identity”) and a separate assessment of theories about what leadership is, may help in disentangling Engle and Lord’s finding. As alluded to in the previous two paragraphs, a major limitation of Engle and Lord’s (1997) study is their measurement of leadership self-schemas as the self-rated importance of traits associated with leadership, but not leadership itself. In no place is “leadership” mentioned in their assessment of leadership self-schemas. Individuals rated themselves on a series of traits and behaviors, but not directly on their self-concepts as leaders. If leadership were a highly emotionally charged or embarrassing topic, implicit assessment might be a particularly useful approach to avoid strong social desirability effects, but leadership is not a particularly sensitive topic and further, it does not appear that the items used to assess leadership self-schemas are any less socially desirable than if leadership self-schemas were more directly measured. Measuring self-assessments of traits associated with leadership is only a partial and indirect measure of self-views about leadership and is conceptually murky. A conceptually tidier way to measure leadership self-views is to directly ask about selfperceptions of leadership. Do individuals see themselves and/or identify as leaders? This direct approach to measuring self-concept is widely adopted in other subject areas (e.g., Crocker, Luhtanen, Cooper, & Bouvrette, 2003; Cross & Markus, 1994) and is used to assess leadership self-view in this dissertation. Before proceeding, it should be noted that Susan Komives and colleagues have also been interested in the general topic of leadership identity (e.g., Komives, Casper, Longerbeam, Mainella, & Osteen, 2004). Their research, which is based in the education literature, has taken a qualitative approach to understanding the development of leader identity by studying a small number of college students who had previously been identified as effective relational leaders. They argue that positive (particularly group) experiences, and a motivation to grow, learn, and make friends contributed to a relational leadership identity, although their results should be interpreted with caution due to methodological problems.
9 One methodological issue in the work of Komives and colleagues is that the only students selected for participation in this study were identified as effective relational leaders. In other words, they only chose to study effective leaders, and therefore did not have a comparison group. In addition, although interviews attempted to re-trace the development of each student’s leadership identity post-hoc, individuals often have difficulty reporting accurately about their own past ways of thinking. Further, the extremely small sample of students in the study warrants caution in interpretation. Finally, leadership identity was never succinctly defined nor measured. In order to understand the development of leadership identity (a useful goal of future research), there must be a reliable and meaningful way of measuring leadership identity. A reliable and meaningfully valid measure of leadership selfidentity is a goal of this study, and is presented in detail in Chapter 5. Leadership Self-Identity In order to avoid confusion with the existing indirect measure of leadership selfschema, the primary construct of interest in this research is termed “leadership self-identity”. This name also describes more succinctly the intent of the construct; the self-identity of an individual in the leadership domain. Consistent with other research in the self-schema literature (e.g., Pelham & Swann, 1989), leadership self-identity is conceptualized and assessed along three related sub-dimensions: self-descriptiveness, certainty, and personal importance of that self-identity.
10 CHAPTER 2 Origins (and/or Correlates) of Leadership Self-Identity It is impossible to determine the true origins of an individual’s leadership selfidentity in real-world settings (given that people cannot be raised in a completely controlled environment), yet there are conceptual reasons to assert that certain prior factors are likely to lead to an individual’s level of leadership self-identity. At the very least, these factors can be considered correlates of leadership self-identity, and, more likely, should be considered to act in a bi-directional causal fashion with leadership self-identity. Leadership self-identity is believed to be correlated with (and influenced by) three broad sources: individual differences in past leadership experiences, personality, and the self-matching of implicit leadership traits to self-assessments on those traits.
Experience According to early work on self-schemas (Markus, 1977), an individual’s schema in a given domain is the result of the self-discovered stability in his or her own social behavior. In this view, past experience influences self-schema in a unidirectional manner. Such a simplistic view of rational behavioral pattern recognition is not tenable given the widely accepted view that people are highly interpretive beings with limited processing capability (Fiske & Taylor, 1991; Srull & Wyer, 1993) and that self-schemas also influence even early experience, but it is certain that behaviors and experience do contribute to views of the self (e.g., Cross & Markus, 1994). In this way, leadership experiences are likely to lead to increased leadership self-identity. In part, the frequent activation of particular aspects of the self-concept should increase the strength, importance, and certainty of a particular self-view. Leadership experience can provide a powerful basis for seeing the self as a leader. Yet not all experiences are equal (Fiedler, 1995) and experience is a multidimensional phenomenon (Quinones, Ford, & Teachout, 1995; Tesluk & Jacobs, 1998). In conceptualizing and measuring experience, quantitative assessments such as tenure in a position or number of times a task was completed have predominated both literature and practice (Tesluk & Jacobs, 1998). Indeed, quantitative assessments are the most straight-forward to conceptualize and assess, but only provide a partial picture of an individual’s experience because they fail to take into account qualitative aspects of experience (Quinones, Ford, and Teachout, 1995). For example, experience with a difficult and complex task is likely to provide more of an opportunity for development and thus be a more beneficial experience than simple tasks. Experience measures which assess the number of times a task has been performed or number of years in a given position might equate the experience of two individuals who, in reality, had qualitatively different experiences.
11 In addition to a qualitative experience component, Quinones et al. (1995) argue that there is an interaction component to experience (which Tesluk & Jacobs, 1998, refer to as “density”), such that intensive, challenging experiences (high quality) over a substantial period of time (high quantitative component) are the most developmental. Although empirically untested as a distinct dimension, the importance of experience density is conceptually compelling. In sum, care must be taken to carefully consider that different components of experience may uniquely contribute to one’s self-schema. In the leadership domain, similar differences in quality, quantity and density of experience are likely to affect the “total package” of experience, and thus the extent to which individuals see themselves as leaders. Specifically, leadership experiences that involve complex tasks and difficulties in a variety of settings over time may lead to a consistent and stronger view of the self as a leader. Hypothesis 1: Quantity, quality, and density of previous leadership experiences are positively related to leadership self-identity. The development of the multidimensional leader experience measure for this study is detailed in Chapter 5 (Methods).
Personality Correlates of Leadership Self-Identity In addition to experience creating the soil from which a leadership self-identity can develop and be strengthened, several personality variables are likely to be associated with leadership self-identity. These factors include core self-evaluation (CSE; Judge, Erez, Bono, & Thoresen, 2003), motivation to lead (Chan & Drasgow, 2001), and self-monitoring (Snyder, 1974). Core Self-Evaluations First proposed by Judge, Locke, & Durham (1997) as a way to understand the dispositional component of job satisfaction, core self-evaluation is conceptualized as a broad trait that represents the fundamental, enduring, and basic appraisal individuals make about themselves. In essence, one’s core self-evaluation (CSE) is the overall self-appraisal of confidence, worth, and agency, and is the latent higher-order factor underlying four commonly studied personality traits: self-esteem, generalized self-efficacy, locus of control, and neuroticism. The fundamental nature of this trait means it is considered to be a source, or central, trait (Cattell, 1965; Rokeach, 1972) that is the underlying cause of more peripheral traits, beliefs, evaluations, and actions. Although CSE has not been assessed as a predictor or correlate of leadership or leadership self-perceptions, it is not difficult to infer that self-confident, positive, agentic
12 individuals (as compared to those who are not) would be more likely to see themselves as leaders. In reviews of the leadership literature, these personality traits are at least related to leadership, if not directly to leadership self-identity. The personality profile of a leader usually includes traits such as self-confidence and persistence (House & Aditya, 1997; Kirkpatrick & Locke, 1991; Stogdill, 1948; 1974), traits that are partial indicators of the CSE construct. Other indirect evidence that CSE is likely to be related to leadership self-identity comes from a recent meta-analysis (Judge, Bono, Ilies, & Gerhardt, 2002), which found that the CSE component trait of emotional stability (inverse of neuroticism) is related to leadership emergence and effectiveness (r = .24). Yet, while personality is related to leadership, leadership emergence, and leadership effectiveness, care must be taken in making the leap to necessarily inferring that personality is thus also related to leadership self-identity because the link between leadership self-identity and leadership outcomes (effectiveness, emergence) is unknown. Fortunately, Engle and Lord (1997) provide more direct evidence that core self-evaluations might be related to leadership self-identity. Engle and Lord found that leadership self-schemas were positively related to positive affectivity and negatively related to negative affectivity in both supervisor and subordinate samples. One of the defining characteristics of high CSE individuals is that they have a high tendency to be positive and a low tendency to be negative (Judge, Erez, Bono, & Thoresen, 2003; Judge, Thoresen, Pucik, & Welbourne, 1999). Further, neuroticism (one of the component traits of CSE) is often viewed as being highly similar to negative affectivity (Costa & McRae, 1980; Watson & Tellegen, 1985), and extraversion (another of the component traits of CSE) is related to positive affectivity (Watson, Clark, & Tellegen, 1988). As a result of the conceptual overlap between affectivity and core self-evaluations, it would be expected, by extension from the Engle and Lord finding, that individuals with a high core self-evaluation would be more likely to see themselves as a leader. Hypothesis 2: A higher core self-evaluation is related to stronger leadership self-identity. Motivation to Lead Chan (2000; Chan & Drasgow, 2001) recently proposed an individual differences construct that affects a leader’s decisions to assume leadership training, leadership roles and responsibilities, and persistence and intensity in leadership effort. This higher-order construct, called motivation to lead (MTL), is composed of three separate but highly related factors: affective-identity MTL (like to lead others), noncalculative MTL (the person’s calculative beliefs about the costs of taking on leadership, where less calculative individuals are less avoidant of leadership roles because they do not calculate the burden of leadership), and social-normative MTL (lead out of a sense of duty or responsibility). Motivation to lead was found to be predictive of assessment center-, peer-, and superior-rated measures of leadership potential. Further, MTL predicted these behavioral measures of leadership potential beyond cognitive ability, individualism/collectivism, and Big 5 measures of personality (Chan & Drasgow, 2001), suggesting that motivation to lead is an important construct in the leadership domain.
13 With regard to leadership self-identity, there is reason to believe that MTL is highly related to leadership self-identity. In the section on self-schemas, it was noted that selfschemas frame how we see the world, what activities we are likely to engage in, and how much persistence we are likely to exhibit when faced with difficulties. In essence, selfschemas can affect our motivation by providing us with an atunement to issues related to a given domain’s self-schema, and a desire to perform in that given domain (Markus, 1977). In the leadership domain, then, it is expected that a strong leadership self-identity might affect the motivation to lead. More specifically, it would be expected that seeing oneself as a leader is related to liking to lead (affective-identity MTL) and leading out of a sense of duty (social-normative MTL). The relationship of leadership self-identity to noncalculative MTL is, at least conceptually, a bit more tenuous. Noncalculative MTL presumes that people are more likely to lead if they do not calculate the costs and burdens that are associated with taking a leadership position. Items in the noncalculative MTL scale tap into a focus on the group, rather than the individual. Not surprisingly, noncalculative MTL is positively correlated with two measures of collectivism (Chan & Drasgow, 2001). Although noncalculative MTL was found to be positively related to ratings of leadership potential (Chan & Drasgow, 2001), and thus by extension it might be expected that it is also positively related to leadership self-identity, significant logic exists to assert that noncalculative MTL is associated with lower leadership self-identity. Those who see themselves as leaders (strong leadership self-identity) are likely to think about leadership more often (based on the self-schema literature), including personal costs of leadership. In addition, a strong leadership self-identity might be related to a focus on individual goals, rather than on collective goals which are characteristic of noncalculative MTL, particularly when one holds a view of leadership as an individual, top-down phenomenon (discussed further in the next section). Taken as a whole, there is a greater conceptual rationale to assert that noncalculative MTL is related to lower, rather than higher, leadership self-identity. Hypothesis 3a: Higher “affective-identity” motivation to lead is positively related to leadership self-identity. Hypothesis 3b: Higher “social-normative” motivation to lead is positively related to leadership self-identity. Hypothesis 3c: Higher “noncalculative” motivation to lead is negatively related to leadership self-identity. Self-Monitoring In addition to core self-evaluations and motivation to lead being related to leadership self-identity, the individual difference variable of self-monitoring is also likely to be related to leadership self-identity. First introduced by Snyder (1974), self-monitoring is defined as the extent to which individuals observe, regulate, and control how their “self” is presented to others. Individuals who are high self-monitors care about others’ reactions to them, are
14 perceptive about others’ reactions to them, and are able to determine (and follow through with) behaviors appropriate to a given situation (Gangestad & Snyder, 2000). At the core of the self-monitoring construct is the construction of positive social appearances. There exists some complexity, however, in determining the effect of self-monitoring on leadership self-identity. Throughout this document, I have considered the three components of leadership self-identity together (i.e., the hypotheses related to leadership selfidentity have considered leadership self-identity as a unidimensional construct, without regard to the nuances related to the unique descriptiveness, importance, and certainty dimensions). They have been considered together because there is no reason to believe that each of the three self-identity sub-components should be differentially related to the other variables discussed up until this point. However, for self-monitoring, it is plausible that the descriptiveness, importance, and certainty components of leadership self-identity may operate differentially. Descriptiveness. In a recent meta-analysis of self-monitoring in the context of organizationally relevant variables, high self-monitors were found to be better job performers, more involved in their jobs, and more effective (and emergent) leaders (Day, Schleicher, Unckless, & Hiller, 2002). In the domain of leadership, the positive impressions among others that high self-monitors engender, the rewarding interactions that high selfmonitors have with others, and their ability to discern and respond to others’ needs4 are likely to lead to leadership emergence and leadership effectiveness (see Day, et al., 2002). What is less clear from this evidence is the extent to which high self-monitors will see leadership self-identity items as descriptive of themselves. A recent self-monitoring construct-analysis paper by Gangestad and Snyder (2000) sheds some light on this issue by postulating that high self-monitors are likely motivated by the appearance of being “high-status”. High self-monitors see themselves as high status, and want others to see them as high status. Given that high status is associated with leadership (Hogan & Hogan, 2002), it is expected that high self-monitors, who see themselves as high status and want others to see them as high status, also see leadership items as descriptive of themselves. Hypothesis 4a: High self-monitors will rate leadership self-identity items as more descriptive of themselves. Importance and certainty. Contrary to the direction of the descriptiveness dimension prediction of leadership identity, it is plausible that high self-monitors are less certain of their ratings of leadership self-identity and that their ratings are less important to their identity. It has been demonstrated that high self-monitors have more difficulty thinking about themselves and that self-knowledge and attitudes of high self-monitors are less accessible than those of low self-monitors (DeBono, Green, Shair, & Benson, 1995; Kardes, Sanbonmatsu, Voss, & Fazio, 1986; Snyder & Cantor, 1980). This is consistent with the 4
Gangestad & Snyder (2000) have recently argued that attention and responsiveness to others may not be as closely linked to self-monitoring as was once believed, though an analysis of self-monitoring measures, particularly the Lennox & Wolfe (1984) suggests that indeed it is a central component.
15 notion that the self-identity of high self-monitors is more malleable. Further, because high self-monitors easily change their self presentation, it is likely that a particular view of the self is not held as highly important. It would thus be expected that high self-monitors are less certain of their self-assessments in the leadership domain and that these self-assessments are less important to their view of self. Hypothesis 4b: High self-monitors will be less certain of their leadership selfidentity than low self-monitors. Hypothesis 4c: High self-monitors will rate leadership self-identity items as having less self-importance than low self-monitors.
Pattern Matching: Matching Self-Traits to Traits of Leader In addition to personality and experience being related to leadership self-identity, a third factor is likely to be influential in seeing oneself as a leader: pattern matching. One of the ways that individuals identify others as leaders is through observing behaviors and traits of target individuals and matching those to implicit theories about what a leader should be like (Bass & Avolio, 1989; Lord, Foti, & DeVader, 1984; Lord & Maher, 1991). If a target’s traits and behaviors match the traits and behaviors that are part of an observer’s implicit leadership theory (ILT), then the target is perceived as a leader. The same argument can be made for self-observation of our own traits and behavior; individuals come to identify themselves as leaders by assessing the match between perceived behaviors and traits of the self and perceived behaviors and traits in the individual’s implicit leadership theory. This pattern matching is what Engle and Lord (1997) referred to as leadership self-schema (their proxy for leadership self-identity). I argue that pattern matching should be conceptually related to a direct measure of leadership self-identity, although the two concepts are distinct. More specifically, it is possible that the pattern matching approach may lead to a leadership self-identity; or, individuals who have a leadership self-identity may come to see themselves as having leadership traits. In either causal case (or in a bidirectional case), a self-assessment that one possesses self-traits that are also widely held leadership traits should be related to a leadership self-identity. There is empirical evidence that leadership prototype traits and behaviors are related to self-assessments of traits and behaviors (Engle and Lord, 1997; Keller, 1999), though their relationship to a more direct measure of leadership self-identity has not been established. Thus: Hypothesis 5: Self-assessment of leadership traits and behaviors (leadership self-schema) is related to leadership self-identity.
16 CHAPTER 3 Individual Theories about the Nature of Leadership In addition to personality and experience correlates of leadership self-identity, it is also believed that there are important outcomes associated with leadership self-identity. But before criterion variables can be meaningfully discussed, it must first be recognized that important outcomes are likely not only the result of whether (or not) we see ourselves as leaders, but also what we believe leadership to be. People have different beliefs about the nature of (effective) leadership, and these perceptions help to dictate leadership style and outcomes (Cohen & March, 1974). For example, some people see leadership as being predominantly about power, while others see leadership as being about guiding followers. These (and other) differences are likely to have significant consequences for how leadership self-identity is expressed. This chapter explores three fundamental approaches to understanding what leadership is, and discusses these orientations in relation to how we see ourselves as connected/disconnected from other people. Orientation Toward Leadership Research beginning with Lord (Lord, Foti, & DeVader, 1984) has demonstrated that both leaders and subordinates (indeed, all individuals) hold implicit theories about leaders (e.g., Engle & Lord, 1997; Epitropaki & Martin, 2004; Offerman, Kennedy, & Wirtz, 1994), and that these implicit theories provide an individual lens through which leadership is perceived and enacted. At the societal level, the well-known Project GLOBE measured implicit perceptions about the nature of effective leaders by aggregating individual responses of 17,000 managers from 62 different societies, and found traits, skills, and styles that were universally shared across societies, as well as traits that were endorsed only in certain cultures or sets of cultures (den Hartog, House, Hanges, Ruiz-Quintanilla, Dorfman, et al., 1999). In both individual and aggregate level research, this implicit leadership theory approach argues that, to a large degree, leadership is in the eye of the beholder rather than being simply an objective phenomenon devoid of individual (and cultural) interpretation. As such, an understanding of leadership requires an examination of the assumptions and theories that individuals (or societies) hold about effective leaders. At the same time that existing research on implicit leadership theories has served to further advance our understanding of the perceptual nature of leadership, measurement of implicit theories has not closely matched their conceptualization. The main problem with existing measures of implicit leadership theories is that while they conceptually claim to measure leadership theories, they actually assess people’s perceptions of traits associated with leadership, rather than leadership behaviors or, more importantly, one’s actual theoretical perspective on what leadership is. For example, in a validation study of implicit leadership by Epitropaki & Martin (2004; based on the commonly used measure of Offerman, Kennedy, & Wirtz, 1994), all of the items tap trait clusters that are associated (positively or negatively) with leadership. An alternative possibility for understanding individual’s theories about leadership is to assess their conceptualizations and orientations about what leadership is, rather than personality traits that are associated with leaders.
17 In exploring different perspectives (or theories) on what leadership is, Drath (2001) proposes that there are three fundamental orientations or three qualitatively different worldviews about leadership: personal dominance, interpersonal influence, and relational dialogue. These principles, or theories, are taken for granted ideas about the nature of leadership that guide further thought and action about leadership. They are ways of seeing and understanding leadership. Using Drath’s writing as a base, a description of the three principles of leadership and their interrelations and implications are explored in the following paragraphs. Personal dominance. The first approach, which considers leadership from a personal dominance perspective, views leadership as something that people possess as a result of their position or individual traits (or both). In this view, leaders influence followers in a unidirectional fashion, from leader to follower. According to this approach, leaders are born (and not made), and are fundamentally different from followers. The personal dominance approach to leadership is not synonymous with domination, although control and power are often common themes in this approach to leadership (Drath, 2001). The core of this approach is that a defined leader is the source of leadership and followers are receivers of leadership. Throughout history, this has been the predominant way in which leadership has been understood. Interpersonal influence. The second approach to leadership, the interpersonal influence approach, emphasizes leadership as a negotiated process of influence, rather than simply being about power. In this orientation toward leadership, leadership is not simply authoritative, top-down leadership; leaders must take into account other perspectives. In cases where a formal leader is not predetermined, leadership involves individuals actively engaging in the process of negotiating leadership, with the leader emerging as the most influential person in the group (although not necessarily the person who is in the highest position, see Follett, 1924). In cases where there is a defined position of “leader”, effective leadership is seen more as a process of influence, rather than coercion and power utilization. Transformational leadership, which includes the components of idealized influence, inspirational motivation, intellectual stimulation, and individualized consideration (Bass & Avolio, 1994) is probably best thought of as fitting into the interpersonal influence approach to leadership. This approach does not preclude the use of personal dominance in some situations, but rather focuses generally on the principal of influence by the leader as the most effective method of enacting leadership. Shared/Relational dialogue approach. Finally, the relational dialogue approach to leadership views leadership from a drastically different position. According to this way of seeing and understanding leadership, it is a shared process of meaning-making. This orientation is based on the idea that the difficult demands of modern life are too much for a single leader to handle (see Pearce & Conger, 2003), and the focus is on the process of leadership, which inherently includes “follower” involvement. Leadership is seen as a property of the group, rather than one individual, and is understood to emerge as a result of collaborative interaction. While thought of somewhat abstractly by Drath (2001), a concrete approach to the relational dialogue principal of leadership sees leadership tasks as a shared responsibility of all group members.
18 Far from being mutually exclusive ways of thinking about leadership, the three approaches may be thought of as composed of concentric rings, with personal dominance in the center, interpersonal influence surrounding personal dominance, and relational dialogue surrounding interpersonal influence (see Figure 1). In Drath’s thinking, a theory of interpersonal influence does not deny the possibility of personal dominance in leadership, but rather it includes both personal dominance and interpersonal influence. Similarly, a relational dialogue approach to leadership sees leadership as a collective process, but acknowledges the existence and relevance of personal dominance and interpersonal influence in leadership. On the other hand, according to Drath, a lower-order approach to leadership is unable to understand those above it. For example, a personal dominance approach is unable to understand leadership as interpersonal influence or relational dialogue. In essence, people who hold a wider approach to leadership are able to see that the narrower approach is simply an approach to leadership, rather than it being the only approach. Kegan (1994), in his influential book In Over Our Heads, describes this general phenomenon as being able to “hold as object” a narrower approach, rather than simply seeing it “as subject”. Wider approaches can see and understand narrower orientations, but narrower orientations are not able to see wider approaches at all, or they are seen as something different and irrelevant. In essence, wider leadership orientations represent increased cognitive complexity (Streufert & Nogami, 1989) around leadership. While these ideas about relationships between levels are conceptually intriguing, the extent to which they hold up to empirical scrutiny is uncertain (and a corollary goal of the present work).
Relational Dialogue Interpersonal Influence
Personal Dominance
Figure 1. The three orientations toward, or world views of, leadership (based on Drath, 2001).
19
Based largely on Drath’s conceptualization and description of the three dimensions of orientation toward leadership, a set of items was developed to capture the essence of each dimension. A full description of the scale development process is detailed in Chapter 5. Level of Self-Construal (Individualism and Collectivism) An individual’s orientation toward leadership is not simply the result of random variation; it is believed that individuals reliably differ in their leadership orientations according to other factors. One factor that is likely to have the most direct implication for leadership orientation is the extent to which the self is seen as connected/disconnected from others, or what Markus & Kitayama (1991) have termed “level of self-construal”. Much of the work in the self-construal domain has focused on differences between cultures along an individualism/collectivism dimension (Hofstede, 1980; Triandis, 1989)5. At its most basic level, collectivists define themselves in terms of roles, relationships, and group memberships, whereas individualists see the self as an entity independent from others (Cross, Bacon, & Morris, 2000). Although early work based on Hofstede’s global analysis from IBM employees in dozens of different countries conceptualized individualism and collectivism as opposite ends of a single continuum, there is empirical evidence to suggest that individualism and collectivism are separate dimensions, and that both are multidimensional (Oyserman, Coon, & Kemmelmeier, 2002; Singelis, Triandis, Bhawuk, & Gelfand, 1995). For example, measures of individualism often differentially predict from measures of collectivism, and collectivism in one part of the world (i.e., Korea) may look entirely different from collectivism in another part (i.e., Isreal; Triandis, 2001). In seeking to address these issues, Singelis and colleagues (Singelis et al., 1995) analyzed existing measures of individualism and collectivism and developed a four-factor, medium bandwidth measure that assesses both horizontal and vertical aspects of individualism and collectivism. The two types of collectivism, horizontal and vertical, both place similar value on interdependence with others, but differ in the extent to which equality is expected. Most simply, horizontal collectivism (HC) is a pattern in which the individual sees group members as similar to him-/her- self and sees the self as a component of the group. In this view, equality is highly valued. Vertical collectivism (VC) is a pattern in which the individual sees the self as part of a group, but sees status differences within the group as inevitable and unobjectionable. Although interdependence is valued, equality is not expected. Similar to collectivism, the two components of individualism are bifurcated into horizontal and collective components. Horizontal individualism (HI) is the pattern where the self is considered relatively autonomous and independent, but equality is still valued, whereas vertical individualism (VI) sees people as being different and unequal. The most
5
Individualism and collectivism are sometimes also referred to as independence and interdependence, respectively (e.g., Heine & Renshaw, 2002; Gabriel & Gardner, 1999; Lee, Aaker, & Gardner, 2000)
20 striking difference between these two components of individualism is that the VI pattern is competitive and self-reliant, whereas the HI pattern does not emphasize these qualities. Cross-cultural differences along these four dimensions have been examined and noted in numerous studies, but it is also important to understand that there are differences within cultures (see Oyserman et al., 2002, for a comprehensive meta-analysis of individualism and collectivism between and within cultures). Cultures are not monochromatic; individuals within cultures differ in their self-construals, and these within-culture individual differences can and have been reliably studied (Brockner, Chen, Mannix, Leung, & Skarlicki, 2000; Gabriel & Gardner, 1999; Singelis, 1994; Wagner, 2002). The present research adopts this same view that the four individualism and collectivism dimensions are meaningful and reliable differentiators of individuals within cultures. But what are the implications for leadership orientation? Given that views of the self have an effect on cognitions, attitudes, and behaviors (Markus & Kitayama, 1991; Singelis, 1994), self-construals are likely related to an individual’s views of what leadership is – their orientation toward leadership. In the following paragraph, several specific hypotheses are developed. Individualism Dimensions Individuals high on vertical individualism, who emphasize the self and also emphasize competitiveness and power differences (Singelis et al., 1995), will have an orientation toward leadership that is high on personal dominance, which views leadership as being about power, authority, formal position, and dominance. At the same time, vertical individualists, who emphasize both hierarchy and status (e.g., Triandis & Gelfand, 1998), are less likely to appreciate an approach to leadership that emphasizes leadership as being a property of the group and the shared domain of multiple group members. Thus we would expect that vertical individualism is negatively related to shared orientation toward leadership. Hypothesis 6a: Vertical individualism and dominance orientation toward leadership are positively related. Hypothesis 6b: Vertical individualism and shared orientation toward leadership are negatively related. The picture of an individual high on horizontal individualism is somewhat different from the high vertical individualist. Horizontal individualists (i.e., those high on horizontal individualism) similarly emphasize the importance of the self and agency, however these individuals value equality and are not particularly interested in being distinguished or having high status (Triandis & Gelfand, 1998). The shared dimension of leadership orientation, which views the possibility that leadership can be a phenomenon of multiple group members, is likely to be positively related to seeing the self as autonomous but equal in status with others.
21 Hypothesis 6c: Horizontal individualism and shared orientation toward leadership are positively related. Horizontal individualism is also likely to be related to an influence/development approach to leadership. There are two main considerations that lead to an hypothesis about the relationship between these two constructs. First, the influence/development leadership orientation emphasizes leadership as being about influence, and not status (Singelis et al., 1995); notions that are likely to appeal to the horizontal individualist who sees the importance of the self, but does not particularly value high status. In addition, influence/ development views leadership as something that can be developed, and this individual development requires a degree of autonomy and agency that horizontal individualism promotes. Thus: Hypothesis 6d: Horizontal individualism and development/influence Orientation Toward Leadership are positively related. Collectivism Dimensions As mentioned previously, both the vertical and horizontal dimensions of collectivism are concerned with the extent to which the self is defined by the group. These two dimensions differ, however, in the extent to which equality and status are valued, which in turn may have important implications for leadership orientations. Horizontal collectivism views the self as part of a group and views group members as similar to each other (Singelis et al., 1995). Because of the emphasis on groups, group belongingness, and an emphasis on equality (and not hierarchy), the horizontal collectivist, then, might be more likely to accept the notion of leadership as a shared process that group members can be involved in, as well as acceptance of the view that leadership can be a property of the group. Conversely, horizontal collectivists are less likely to endorse a top-down, hierarchical approach to leadership that is emphasized in the dominance leadership orientation. Further, Project GLOBE research (Gelfand, Bhawuk, Nishii, & Bechtold, 2004) at the societal level found support for the notion that collectivistic cultures are more endorsing of interdependence, collaboration, and leaders who are nurturant, relational, and emphasize the group. Thus: Hypothesis 6e: Horizontal collectivism and shared orientation toward leadership are positively related. Hypothesis 6f: Horizontal collectivism and dominance orientation toward leadership are negatively related. Just as individualism and collectivism are unique dimensions (as opposed to being opposite ends of a continuum; Oyserman, Coon, & Kemmelmeier, 2002), leadership orientations are seen as being unique factors that are not diametrically opposed. It is possible that both a shared leadership orientation and a dominance approach may be endorsed simultaneously in an individual who recognizes the existence and value of both approaches
22 to leadership. At the same time that a vertical collectivist values the collective, he or she is also accepting of status differences between members (Triandis & Gelfand, 1998) and is willing to submit to authority (Triandis, 2001). The emphasis on the collective could allow for a higher shared orientation toward leadership, and thus be there may be a positive relationship with shared orientation and vertical collectivism. Simultaneously, the acceptance of status differences may allow for a view of leadership that endorses formal positions and authority, such as is the case in the dominance leadership orientation. Based on these arguments: Hypothesis 6g: Vertical collectivism and shared orientation toward leadership are positively related. Hypothesis 6h: Vertical collectivism and dominance orientation toward leadership are positively related. In addition to self-construals playing a role in theories about leadership, it is also expected that there is a general tendency of formally appointed leaders to have more independent self-construals than subordinates. In a conceptual paper on power and selfconstruals, Lee and Tiedens (2001) argue that people with legitimate (position) power are likely to have more individualistic and less collectivistic self-construals. They hypothesize, based on the unstated assumption that self-construals can change when an individual takes on a position of power, that power is related to independent self-construals due to powerful individuals’ decreased reliance on others for resources. Conversely, those with less power are more reliant on powerful individuals for resources, and thus would be expected to have self-construals that are more collectivistic and less individualistic. Another possible explanation, leading to the same hypothesis, is that individualistic people are more likely to achieve positions of leadership than collectivists. Collectivists, who see the self as part of the group, may be less likely to aspire to, and be selected for, a position of leadership. Leadership requires sometimes unpopular decisions, and leadership can be a lonely enterprise (Kets de Vries, 2001); individuals high in collectivism (and low in individualism) may seek to maintain a view of themselves as part of the group and thus behave in ways that are less suitable for selection into leadership positions. Based on these two possible explanations, which lead to the same predicted result, it is hypothesized that: Hypothesis 7: Supervisors are more individualistic and less collectivistic than their followers.
23 CHAPTER 4 Outcomes of Leadership Self-Identity and Orientation Toward Leadership An individual’s leadership self-identity and his/her orientation toward leadership are believed to be related to a number of outcome variables, including: leader-member relationship quality, behavioral repertoire, interest in leadership development, and leadership potential. Theoretically, it is believed that the relationship is often a moderated one; orientation toward leadership moderates the link between leadership self-identity and these outcome variables. This chapter explicitly details the hypothesized links between both leadership self-identity, orientation toward leadership, and leadership outcomes. Leader-Member Relationships The development of Leader-Member Exchange (LMX) theory (see Graen & UhlBien, 1995, for an overview) was a breakthrough in the study of leadership because it began a focus on the quality of the relationship between a leader and “member”. Indeed, the overall quality of the relationship between supervisors and subordinates is an important consideration: it is related to job performance, satisfaction, commitment to the organization, and turnover intentions (Gerstner & Day, 1997). But what predicts positive supervisorsubordinate relationships, which are an important outcome variable in their own right? One factor may be the extent to which the positional leader (or supervisor) sees him or herself as a leader. As mentioned previously, one of the findings of the Engle and Lord (1997) study was that supervisors with stronger leadership self-schemas had poorer quality relationships with their subordinates. The suggestion offered for this somewhat unexpected finding was that leaders who think of themselves as leaders are stricter judges because they are referencing the self when they evaluate others, and the self is known to provide a more stringent referent for evaluating others (Dunning & Hayes, 1996). In turn, being a stricter judge of others also leads to lower-quality relationships. Consistent with the related finding of Engle and Lord, I would expect that there would be an overall effect for leadership self-identity, such that higher leadership self-identity is related to lower quality relationships (as rated by the supervisor). Hypothesis 8: Leadership self-identity is negatively related to supervisorrated leader-member exchange (LMX) quality. Leadership Potential According to Engle and Lord’s argument (and the literature on self-schemas), it is also expected that having a strong self-schema in a given domain is related to lower evaluations of others in that domain. Specifically, a strong supervisory leadership selfidentity would be expected to be related to lower evaluations of subordinate leadership potential. This important possibility was not assessed in Engle and Lord’s study. However,
24 an assessment might provide some insight into the veracity of their explanation for why leadership self-perceptions are negatively related to LMX quality. While not a definitive test (because potential real differences in subordinate leadership potential across different supervisors cannot be controlled), it is a starting point. Hypothesis 9: Supervisors with a strong leadership self-identity will give lower ratings of subordinate leadership potential than supervisors with a weaker leadership self-identity. Orientation Toward Leadership as a Moderator But is leadership self-identity always related to being a more stringent judge of others in a given domain and to lower relationship quality? According to Engle and Lord (1997), the answer would appear to be “yes”, yet it seems plausible that one’s orientation toward leadership may provide a moderating force. In other words, it is important to understand whether individuals see themselves as leaders (leadership self-identity), and also understand what they believe leadership to be (orientation toward leadership). Some individuals may see themselves as leaders and still see leadership as a process that includes subordinates, rather than simply a hierarchical power-play that requires a dominant approach. In a less dominantfocused and more relational-focused approach to leadership, the effects of leadership selfidentity on LMX and ratings of subordinate leadership potential might be moderated. In cases where a supervisor sees him/herself as a leader and thinks that leadership is about one person dominating and controlling subordinates (i.e., a personal dominance approach), relationship quality and downward ratings of leadership potential might be lowest. Hypothesis 10a: Orientation toward leadership moderates the relationship between a supervisor’s leadership self-identity and relationship quality (LMX). Leadership self-identity is more negatively related to LMX ratings when a supervisor’s leadership orientation is highly dominant and low on interpersonal influence and relational dialogue. Hypothesis 10b: Orientation toward leadership moderates the relationship between a supervisor’s leadership self-identity and ratings of subordinate leadership potential. Leadership self-identity is more negatively related to leadership potential when a supervisor’s leadership orientation is highly dominant and low on interpersonal influence and relational dialogue. Behavioral Repertoire Another way to think about leadership outcomes is in terms of the variety of behaviors that a leader is able to engage in, also known as a behavioral repertoire or behavioral complexity. The assumption (and suggestive findings) of this line of research is that behaviorally complex leaders are more effective because they are able to adapt to the requirements of different situations by enacting different behaviors (Zaccaro & Klimoski,
25 2001). Leaders who are only able to utilize one type of behaviors may be effective in certain circumstances, but their overall effectiveness across situations will be diminished because of the requirements of complex and “competing” behaviors in different situations. Based on Quinn’s (1988) competing values framework (Figure 2), behavioral repertoire is most easily thought of as the frequency with which a leader engages in four leadership functions (and behaviors) within the people, adaptive, stability, and task domains (Denison, Hooijberg, & Quinn, 1995). In a study of 334 middle level managers from two different companies, Hooijberg (1996) showed that behavioral repertoire, calculated as the overall mean of the four leadership functions, is highly related to leadership effectiveness (with zero-order correlations from .46, .58, and .64 [peers, superior, and subordinate rated]). Although the study suffered from the potential for same-source bias (effectiveness ratings and behavioral repertoire were assessed by the same individuals), the findings are suggestive that supervisors who are able to more frequently engage in the different (and competing) behaviors were seen as more effective by their subordinates, peers, and supervisors. But what might lead to an increase in behavioral repertoire? Because self-schematics in a given domain are known to spend more time thinking about that domain and their thinking is generally more “complex” and refined, and because thinking is for doing (Fiske, 1992), it is expected that a strong leadership self-identity is related to having more possible leadership behaviors at one’s disposal. Thus, a strong leadership self-identity is expected to be related to increased behavioral complexity (repertoire). Hypothesis 11: Leadership self-identity of supervisors is positively related to subordinate ratings of supervisory behavioral repertoire. In addition to the proposed main effect of leadership self-identity on behavioral repertoire, it is also believed that a supervisor’s orientation toward leadership directly affects subordinate ratings of supervisory behavioral repertoire. Given that a wider leadership orientation (one that includes high scores on the influence and relational dialogue/shared dimensions) is believed to be a type of increased complexity in thinking about leadership, this study is in a unique position to test the possibility that cognitive and behavioral complexity are linked. Again, because thinking is for doing (Fiske, 1992), it is expected that a more complex leadership orientation (e.g., higher scores on the “higher-order” dimensions of Interpersonal Influence and Shared/Relational Dialogue) would be associated with an increase in behaviors available (and performed) by a supervisor. Hypothesis 12: A wider supervisor-rated leadership orientation is positively related to subordinate ratings of supervisory behavioral repertoire6.
6
Orientation toward leadership was not hypothesized to moderate the self-identity – behavior link in this case because of the hypothesized strong and theoretically interesting main effect of implicit orientation on behavioral repertoire.
26
Flexibility = competing values
People Function e.g., mentor, facilitator
Adaptive Function e.g., innovator, broker
Internal
External Stability Function e.g., monitor, coordinator
Task Function e.g., director, producer
Control
Figure 2. Quinn’s Competing Values Framework.
Interest in Leadership Development A consistent finding in the self-schema literature is that individuals with strong selfschemas are better performers (Cross & Markus, 1994) and seek to demonstrate their proficiency in a given domain (Markus, 1977). They pay close attention to, and favor information in a given domain, and interpret events according to their self-schemas. In addition, self-schematics have a higher degree of self-efficacy about their abilities in a given domain and are persistent. Consistent with what is known about self-schematics in general, it is expected that individuals who are self-schematic in the leadership domain (i.e., those who have a leadership self-identity) are more likely to pursue opportunities to increase their abilities in the leadership domain, or show their abilities off (Markus et al., 1990). Regardless of the exact underlying causal mechanism (increased abilities or desire to demonstrate proficiency), I would expect that leadership self-identity is related to a higher degree of involvement in
27 leadership development activities7. In the sense that leadership self-identity is related to future participation (or stated interest) in leadership development, it may be true that “the rich get richer”. Thus, a supervisor with a strong leadership self-identity would be expected to be a more effective leader (unfortunately this was not tested by Engle & Lord). In addition, self-schematics have a desire to be seen by others as they see themselves, so a supervisor with a strong leadership self-identity would have a desire to be seen as a leader, leading him or her to do things to be seen as an effective leader. Thus, it is hypothesized that: Hypothesis 13a: For both supervisors and subordinates, leadership selfidentity is positively related to willingness to develop as a leader. In addition to leadership self-identity playing a role in interest in leadership development, one’s view about leadership is also likely to be important. More specifically, development orientation toward leadership, which partially assesses one’s beliefs about whether leadership can be taught, trained, and developed, is likely to be an important predictor of whether one will be interested in engaging in leadership. If an individual believes that leadership cannot be developed, participating in leadership development would be seen as pointless, and thus I would expect a correspondingly low level of interest in leadership development. Conversely, the belief that leadership can be developed is likely to lead to increased interest in leadership development. Hypothesis 13b: For both supervisors and subordinates, a high development/influence orientation toward leadership is positively related to willingness to develop as a leader. Further to the leadership self-identity – leadership development main effect hypothesis, and consistent with the other outcome hypotheses, orientation toward leadership is seen to be an important moderator. Highly schematic individuals whose orientation toward leadership includes the belief that leadership can be developed (part of the relational leadership approach) are more likely to be willing to develop their own leadership than individuals who believe that leadership is an innate characteristic (low shared and high dominance orientation). Hypothesis 14: Orientation toward leadership moderates the relationship between leadership self-identity and willingness to engage in leadership development. Leadership self-identity is more positively related to leadership development when an individual’s leadership orientation is low on personal dominance and high on development/interpersonal influence and relational dialogue.
7
It is, at first glance, plausible to reason that leadership self-identity may lead to a decreased willingness to participate in leadership development activities if those activities were seen as highly threatening to one’s selfview. However, self-schematics are not easily threatened in their self-schematic domain.
28 Does Leadership Self-Identity Always Matter? – Boundary Conditions Generally, individuals who are schematic in a domain (i.e., have a self-identity in a given domain) perform differently than individuals who are aschematic; however, under certain conditions performance may be similar. For example, Cross and Markus (1994) showed that when external motivation or environmental pressures are strong enough, aschematic performance may be similar to schematic performance. Schematic individuals may differ from aschematics particularly in conditions when a task requires significant structuring by the individual (such as a novel task) or when significant obstacles are present (Kendzierski & Sheffield, 2000). Similar to strong and specific situations acting as substitutes for leadership (Kerr & Jermier, 1977), certain situations may also act as substitutes for leadership self-identity. However, in modern organizations, leadership often takes place in conditions of novelty, ambiguity, and varying degrees of motivation, and thus it would generally be expected (particularly over a number of episodes) that individuals schematic for leadership will behave differently from aschematics. Two Proposed Studies To answer the substantive research questions that were proposed, two studies (one using a student sample and the other an organizational sample) will be used to assess variables related to leadership self-identity and orientation toward leadership. The primary purposes of the first study were to: a) assess theoretical antecedents and correlates of leadership self-identity and leadership orientation and b) determine the factor structure of these newly developed measures. The purposes of the second study were to: a) assess the generalizability of the factor structure findings in an organizational sample and b) assess organizationally relevant outcomes associated with leadership self-identity and leadership orientation in an field setting (see Figure 3).
29
Antecedents
Self-Identity
Outcomes
Level of SelfConstrual (Individualism/ Collectivism)
Supervisor ratings of LMX with subordinates
Previous leadership experience (quantity, quality, density) Core Self-Evaluations
Orientation Toward Leadership
Increased behavioral repertoire of supervisors (subordinate-rated)
Motivation to Lead Self-Monitoring
Leadership Self-Identity
Implicit Leader traits match with selfratings on those traits
Supervisor and subordinate interest in leadership Supervisor ratings of subordinate leadership potential
Student Sample
Medical Center Sample
Figure 3. Overall model of proposed antecedents and outcomes of leadership selfidentity and orientation toward leadership: Study 1 and 2.
30 CHAPTER 5 Development of the Leadership Self-Identity and Orientation Toward Leadership Measures
This chapter describes the development and refinement of the Leadership SelfIdentity measure and the Orientation Toward Leadership measure. Development of the Leadership Self-Identity Scale The existing measure of leadership self-schema developed by Engle and Lord (1997) asks participants to rate themselves on a number of traits associated with leadership, but does not directly assess the extent to which an individual sees him/herself as a leader. Given this problem, and the current interest in assessing self-views of leadership, a measure was developed to more directly assess leadership self-identity. After an examination of the style of items used in the self-schema literature, four items were written (“I am a leader”; “I see myself as a leader”; “If I had to describe myself to others, I would include the word ‘leader’”, and; “I prefer being seen by others as a leader”) to adequately cover the content domain of leadership self-identity (see Figure 4). The last item, “I prefer being seen by others as a leader” is included because of the well-known finding in the self-schema literature that individuals with a strong self-schema actively attempt to have others see them as competent in their self-schematic domain (Cross & Markus, 1994). Consistent with other research in the self-schema literature (e.g., Pelham & Swann, 1989), all (four) items were assessed for their self-descriptiveness, the certainty of their selfratings, and the personal importance that each statement holds to the respondent. For the descriptiveness ratings, response anchors ranged from not at all descriptive (1) to extremely descriptive (7). For the certainty ratings, response anchors ranged from not at all certain (1) to extremely certain (7). Personal importance anchors ranged from not at all important to me (1), to extremely important to me (7). At this early stage of research, descriptiveness, importance, and certainty ratings were examined separately, in addition to examination of aggregate overall Leadership Self-Identity scores8.
8
For simplicity, unless otherwise noted, leadership self-identity refers to an overall leadership self-identity score (mean of descriptiveness plus importance plus certainty). In data analysis, both aggregate scores and individual domain scores (e.g., importance ratings only) were examined to better understand the structure and validity of the leadership self-identity construct.
31
Leadership Self-Identity Measure All items answered on 1-7 Likert-type scales. Self-Identity – descriptiveness dimension a) Please rate the extent to which the following statements describe you. (anchors: 1 = not at all descriptive, 7 = extremely descriptive) 1. 2. 3. 4.
I am a leader. I see myself as a leader. If I had to describe myself to others, I would include the word “leader”. I prefer being seen by others as a leader.
Self-Identity – importance dimension b) How important to you is this view of yourself? (anchors: 1 = not at all important, 7 = extremely important) 1. 2. 3. 4.
I am a leader. I see myself as a leader. If I had to describe myself to others, I would include the word “leader”. I prefer being seen by others as a leader.
Self-Identity -- certainty dimension c) How certain are you about each of these statements? (anchors: 1 = not at all certain, 7 = extremely certain) 1. 2. 3. 4.
I am a leader. I see myself as a leader. If I had to describe myself to others, I would include the word “leader”. I prefer being seen by others as a leader.
Figure 4. Leadership self-identity measure.
Development of the Orientation Toward Leadership Measure Given that no existing measure captured the nature of what different individuals think leadership is, an Orientation Toward Leadership measure was developed. Items for this measure, which is theoretically composed of three unique scales, were based largely on Drath’s (2001) work on leadership orientations. The three scales were developed to tap the presumed three leadership orientation dimensions of personal dominance, interpersonal influence, and relational leadership. In addition to Drath’s work, the literatures on social and interpersonal dominance (e.g., Pratto, Sidanius, Stallworth, & Malle, 1994) and
32 authoritarianism (particularly Altemeyer, 1981) informed the Dominance dimension, and transformational leadership (Bass & Avolio, 1994) informed the Interpersonal Influence dimension. Additional conceptual information on these three orientation dimensions can be found in Chapter 3. Item generation. An initial list of 50 items was created by the researcher in an attempt to fully capture the nature of the three leadership dimensions (Appendix A). Eighteen Dominance items, 13 Interpersonal Influence items and 19 Shared/Relational Dialogue items were created. This list of items, along with descriptions of each dimension, was sent to three leadership experts who were not associated with this project (including Bill Drath). The goal of the expert involvement was to receive additional expert opinions about the entire proposed measure, as well as specific items. After receiving expert input and after further discussion, a number of items were removed from each dimension for one or more of the following reasons: the item was not unique to a specific leadership approach, the item may be difficult for participants to interpret, the item could be interpreted as being double-barreled, or a group of items was collectively too cumbersome for efficient administration (e.g., relational dialogue items 710). In addition to deleting items, a number of items were reworded for clarity following discussion and prior to administration of the Q-sort task. Q-sort task. After discussion and paring of items, a total of 37 items were finally retained for further analysis on a Q-sort task. A list of these items (and their intended underlying dimension) is presented in Table 1. The actual Q-sort task given to the 17 expert raters, along with the instructions presented and intended construct for each item, is presented in Appendix B. Expert raters included 8 Ph.D. Industrial/Organizational Psychologists and 9 graduate students in Industrial/Organizational Psychology. Raters were given a description of the dimensions and asked to assign each item to a corresponding conceptual dimension. For judging purposes, in cases where a rater failed to assign a construct to a specific dimension, or in cases where an item was assigned to more than one construct, the response was considered to be improperly assigned to the intended dimension (and thus considered as incorrect). Anderson and Gerbing (1991) have shown that assessments of the substantive validity of items in a Q-sort task can be a useful predictor of items that would be retained in a subsequent confirmatory factor analysis. The two indices that Anderson and Gerbing propose to assess the substantive validity (or item validity, Holden & Jackson, 1979) of items are the: proportion of substantive agreement (psa), and the substantive-validity coefficient (csv). The proportion of substantive agreement is simply “the proportion of respondents who assign an item to its intended construct”: psa = nc/N, where nc represents the number of respondents assigning a measure to its proposed construct and N is the total number of respondents.
33 The substantive validity coefficient reflects the degree to which participants “assign an item to its construct more than to any other construct” (Anderson & Gerbing, 1991, pg. 734), and is calculated as: csv = (nc-no)/N, where no represents the highest number of assignments to a construct other than the construct the item was intended to measure. The critical number of assignments can be obtained from the binomial distribution, such that: P(Y) = (N!/Y!(N-Y)!) * πY(1-π)N-Y where N is the number of respondents, Y is the number of correct responses required, P is the Type I error, and π is .5, the most restrictive probability of assignment of an item to its intended construct (Anderson & Gerbing, 1991). With 17 respondents and a desired alpha of .05, the critical number (m) is 14. One then obtains the critical value of csv as: csv = 2m/N – 1. Substituting 14 for m and 17 for N, the critical value of csv is .64. Substantive agreement and substantive-validity coefficient indices for all 37 original shared leadership items are displayed in Table 1. Of the 37 items, 35 exhibit both satisfactory substantive validity and substantive agreement, suggesting that they be retained in the survey administration. Two items (16 and 26) exhibited a coefficient of substantive validity (csv) below the critical value of .64. These items (“When a group performs well, it is largely because of the leader”, and “Being a leader requires using both rewards and punishment”), however, exhibited satisfactory substantive agreement, suggesting that the items may not be immediately discarded prior to actual survey administration in a full sample. Given the early stage of development of this measure, it was decided to retain both items 16 and 26 for survey administration in Study 1 so that they could be submitted to factor analysis, where a more complete decision could be made. In addition to providing standardized responses, several Q-sort task participants also made comments about items that they thought might be potentially problematic. As a result of these comments, several minor changes were made. Item 7 (from Table 1) was reworded for clarity, from “Leaders should minimize the amount that others question them” to "Leaders should not let others openly question them." This item, in its original wording, did exhibit adequate substantial validity and substantive agreement, but several participants indicated that the syntax of this item could be improved. Item 2 was reworded from “Group members can work together in leadership tasks” to “Group members can work together in performing leadership tasks.” Item 26 was reworded to state “When a group performs well, it is mainly (previous – “largely”) because of the leader.” The referent “they” in item 28 (“One’s formal position determines whether they are a leader”) was changed to “he or she”. Items 32 and 34 (“You’re either a leader, or you’re not” and “You can’t teach leadership”) were reworded for clarity by changing the referent “you” to “people”, so that participants would not be prone to answering the question with reference to the self, but rather with reference to their leadership theories in general. Table 1 contains original (“pre-Q-sort”) wording for all items. Subsequent tables display post-Q-sort wording used in survey administration. Before administering the actual survey to the student and organizational samples, reworded items were presented to an individual who was not involved in the original Q-sort to again check content clarity.
34 Table 1 Results of Item-to-Construct Assignment Task Item #
1 2 3 4 5 6
7
8 9 10 11 12 13 14 15 16 17 18
Item Description (abbreviated)
# Total Intended Construct correct N
psa
highest # wrong
csv
Telling other people what to do is a big part of leadership. Group members can work together in leadership tasks. Leaders order others around.
Dominance
17
17
1.00
0
1.00*
Shared
17
17
1.00
0
1.00*
Dominance
17
17
1.00
0
1.00*
Leadership involves persuading other people. Effective leaders seek different perspectives. Individual people do not possess leadership -- it is a property of the group. Leaders should minimize the amount that others question them. Leadership happens when people collaborate. The leader should be the most influential person in a group. Leadership is a process.
Influenc/Dev.
16
17
0.94
1
0.88*
Influence/Dev.
15
17
0.88
2
0.76*
Shared
17
17
1.00
0
1.00*
Dominance
16
17
0.94
1
0.88*
Shared
17
17
1.00
0
1.00*
Influence/Dev.
14
17
0.82
3
0.65*
Shared
15
17
0.88
2
0.76*
Leadership and power are pretty much the same thing. All group members should participate in leadership. A leader never shows weakness. Leadership is about taking charge of a situation. People can be taught to be more effective leaders. Being a leader requires using both rewards and punishment. Skills and abilities for leadership can be developed. Leadership is a property of the group, not the individual.
Dominance
17
17
1.00
0
1.00*
Shared
17
17
1.00
0
1.00*
Dominance
16
17
0.94
1
0.88*
Dominance
15
17
0.88
2
0.76*
Influence/Dev.
16
17
0.94
1
0.88*
Dominance
12
17
0.71
5
0.41
Influence/Dev.
16
17
0.94
1
0.88*
Shared
17
17
1.00
0
1.00*
35 19 20 21
22
23
24
Leadership is about maintaining control. Leadership involves a group collectively making decisions.
Dominance
17
17
1.00
0
1.00*
Shared
16
17
0.94
1
0.88*
Getting “buy-in” from followers is an important leadership task. Effective leaders influence followers more than followers influence them. Leaders, as people, are fundamentally different from non-leaders. Leadership can be shared.
Influence/Dev.
14
17
0.82
3
0.65*
Influence/Dev.
15
17
0.88
2
0.76*
Dominance
17
17
1.00
0
1.00*
Shared
17
17
1.00
0
1.00*
Leadership is the responsibility Shared 17 17 1.00 0 of everybody in a group. 26 When a group performs well, it Dominance 13 17 0.76 4 is largely because of the leader. 27 Leadership is about using Influence/Dev. 15 17 0.88 2 influence. 28 One’s formal position Dominance 14 17 0.82 3 determines whether they are a leader. 29 Leadership is a two-way street Influence/Dev. 14 17 0.82 3 between leaders and followers. 30 Leaders can acquire skills to Influence/Dev. 17 17 1.00 0 make them more effective. 31 Together, group members Shared 17 17 1.00 0 create leadership. 32 You’re either a leader, or Dominance 17 17 1.00 0 you’re not. 33 Leaders allow followers to Influence/Dev. 14 17 0.82 3 influence them. 34 You can’t teach leadership. Dominance 17 17 1.00 0 35 Leadership is about the group, Shared 17 17 1.00 0 rather than a single leader. 36 Leadership is not possessed by Shared 17 17 1.00 0 any one individual. 37 If you supervise others, you are Dominance 15 17 0.88 2 a leader. * significant at p < .05. psa = proportion of substantive agreement. csv = coefficient of substantive validity. 25
1.00* 0.53 0.76* 0.65* 0.65* 1.00* 1.00* 1.00* 0.65* 1.00* 1.00* 1.00* 0.76*
36 Measure Refinement Refinement of the Leadership Self-Identity Measure One of the goals of this research is to refine and assess the psychometric properties of the Leadership Self-Identity and Orientation Toward Leadership measures. The refinement and assessment process of these measures involves data from both Study 1 (student sample) and 2 (medical center sample), and is presented here, before a specific discussion of each study. In both samples, leadership self-identity was assessed using four items, with each statement rated as to its self-descriptiveness, self-certainty, and self-importance (see Figure 4). A full description of the initial development of the measure can be found in Preliminary Development of the Leadership Self-Identity Scale, earlier in Chapter 5. In total, five different Leadership Self-Identity scores were calculated. These included: (a) a “descriptiveness” score, (b) a “certainty” score, (c) an “importance” score, (d) a “Core Leadership Self-Identity Factor” score, and (e) a grand mean, or “Overall Leadership Self-Identity” score. The first three dimension scores were both theoretically justified and in line with previous research on self-schemas, which use three separate scores for descriptiveness, certainty, and importance dimensions (e.g., Swann & Pelham, 2002). Mean scores on each dimension were calculated, and reliabilities are presented in Chapters 6 (Study 1) and 7 (Study 2). Assessing the factor structure of the Leadership Self-Identity items. Given that the Leadership Self-Identity measure was developed specifically for the present project, a better understanding of the underlying factor structure of the measure was sought. Using data from Study 1, an exploratory factor analysis (using principal axis factoring and varimax rotation) was conducted on all 12 items simultaneously to better understand how they are related. Results revealed two factors with eigenvalues above 1; the first factor explained a substantial amount (55.2%) of variance in the 12 items, and the remaining factor accounted for 13.3% of the variance. Similar results were obtained in a factor analysis of Study 2 data; two factors had eigenvalues above 1. The first factor accounted for 50.8% of the variance in the items, and the second accounted for 19.9%. Table 2 presents factor loadings for data from both Study 1 and Study 2. The first factor in both studies included all items from descriptiveness and importance dimensions; certainty items loaded on the second factor. The eight items comprising the first factor were examined carefully in order to interpret the essence of this factor. The four descriptiveness and four importance items assess the core of leadership self-identity (i.e., seeing the self as a leader). This core of descriptiveness and importance is consistent with some self-schema research that assesses self-schematicity using descriptiveness and importance dimensions and excludes certainty assessment. Thus, this factor is labeled “Core Leadership Self-Identity.” A Core Leadership Self-Identity score was subsequently calculated as the mean of descriptiveness and importance items. Further assessment of this factor in correlation and regression analyses will help to explore the question of whether these dimensions (Descriptiveness and Importance) should be considered separately or together as a central component of leadership self-identity.
37 Certainty items were not included in the Core Leadership Self-Identity score because they failed to load clearly on this factor. A separate certainty factor (Factor 2) was not calculated based on the factor analysis because the items loading on this factor are identical to the certainty dimension items, and therefore these items already formed a calculated dimension (self-certainty). Overall Leadership Self-Identity. Finally, a fifth Leadership Self-Identity score was calculated. The Overall Leadership Self-Identity score was simply the grand mean of the descriptiveness, certainty, and importance scores for each individual. Table 2 Exploratory Factor Analysis of Leadership Self-Identity Items in Studies 1 and 2 Factor Loadings Study 1
Study 2
1
2
1
2
1. Self-Descriptiveness: I am a leader
.644
.310
.599
.303
2. Self-Descriptiveness: I see myself as a leader
.674
.340
.676
.264
.757
.344
.633
.131
.692
.294
.665
.197
5. Self-Certainty: I am a leader
.255
.834
.239
.842
6. Self-Certainty: I see myself as a leader.
.236
.872
.244
.910
.149
.694
.164
.896
.329
.672
.156
.900
9. Self-Importance: I am a leader
.819
.313
.862
.178
10. Self-Importance: I see myself as a leader
.813
.172
.840
.175
11. Self-Importance: when I describe myself to others I include the word “leader”
.348
.052
.763
.089
3. Self-Descriptiveness: describe myself to others as a “leader” 4. Self-Descriptiveness: I prefer being seen by others as a leader
7. Self-Certainty: when I describe myself to others I include the word “leader” 8. Self-Certainty: I prefer being seen by others as a leader
12. Self-Importance: I prefer being seen by .797 .198 .780 .105 others as a leader Note. Principal axis factoring with varimax rotation was used. Bold factor loadings indicate a clear loading on that factor. N (Study 1) = 441, N (Study 2) = 224.
38 Refinement of the Orientation Toward Leadership Scale Item-level data assessment. Thirty-seven Orientation Toward Leadership items were administered in both Study 1 and Study 2. In developing a parsimonious and psychometrically sound Orientation Toward Leadership measure, one of the first tasks was to assess item-level data for problematic characteristics. Descriptive statistics (mean and standard deviation) were examined for each item using both Study 1 and Study 2 data (Appendices I and J). In addition, skew and kurtosis of the items were assessed. None of the items exhibited low variance or unreasonably high or low means (although several items did have means that were fairly high – above 6 on a 7-point scale), and skew and kurtosis values were within the accepted range of +/- 2, suggesting that the items were relatively normally distributed. Thus, none of the 37 items were removed based on the descriptive statistics. Exploratory factor structure. It was believed that three unique factors would emerge (Dominance, Influence/Development, and Shared Leadership), although the precise relationship among these factors was not known. Data from all 37 Orientation Toward Leadership items from Study 1 were subjected to exploratory factor analysis using principal axis factoring and varimax rotation. A total of 9 factors had eigenvalues greater than 1, although factors 5 through 9 contributed very little to the cumulative variance explained (12%) in the items, suggesting that up to four meaningful factors accounted for a large portion of the variance in the items. Another method of determining the number of “significant” factors is an examination of the slope of a plot of factor eigenvalues (Cattell, 1966). An analysis of the factor scree plot (Figure 5) suggested that three factors, and possibly a fourth factor, should be retained in subsequent analysis. Items that failed to load clearly on one of the four factors (with a loading greater than .4 on one factor, and loadings on all other factors of less than .3) were dropped, resulting in 24 items (see Table 3). With one exception (discussed below), the first three factors contained items that corresponded with their theoretical constructs (Shared Leadership, Influence/Development, and Dominance, respectively). It was hypothesized that the influence view of leadership understands leadership as being about development of leaders and group influence. However, in the factor analysis, items on the “Influence” factor in the student sample clustered around only part of this theoretical construct; the development and teaching of leadership. Items assessing influence did not load on any of the factors. Given this factor that emerged, it is not surprising, that the item “You can’t teach leadership” loaded highly and negatively on this second factor, as opposed to loading on the “Dominance” factor. This item, because it logically fit with the other items composing this factor, was retained. In assessing the fourth factor, an examination of items revealed strong similarities with the third factor. Even after careful consideration, it was apparent that the themes of the two factors (dominance view of leadership) were essentially identical and indistinguishable. For example, Item 7 (Leaders should not let others openly question them) loaded highly on the fourth factor, whereas Item 13 (A leader never shows weakness) loaded highly on the third factor. Considering a preponderance of the theoretical and empirical evidence, it was decided to keep the items from the fourth factor.
39
6
5
Eigenvalue
4
3
2
1
0 1
2
3
4
5
6
7
8
9
Factor
Figure 5. Scree plot of eigenvalues by Orientation Toward Leadership factor: Study 1.
40 Table 3 Results from Principal Axis Factoring of Orientation Toward Leadership Items: Study 1 Factor Loadings Item 1. Telling other people what to do is a big part of leadership. 2. Group members can work together in performing leadership tasks. 3. Leaders order other people around. 4. Leadership involves persuading other people. 5. Effective leaders seek different perspectives. 6. Individual people do not possess leadership -- it is a property of the group.
Shared Development Dominance 1
Dominance 2
-.03
.06
.19
.16
.32
.33
-.01
-.06
-.07
-.06
.13
.47
-.09
.03
.07
.06
.14
.21
-.03
-.26
.51
-.11
-.20
.24
.08
-.20
.20
.42
7. Leaders should not let others openly question them. 8. Leadership happens when people collaborate. 9. The leader should be the most influential person in a group. 10. Leadership is a process. 11. Leadership and power are pretty much the same thing.
.61
-.05
-.07
-.04
-.06
.04
.37
.23
.16
.22
.15
-.02
-.07
-.09
.22
.40
12. All group members should participate in leadership.
.52
.14
-.05
.08
-.01
-.03
.49
.29
-.17
.14
.54
-.01
-.01
.77
.09
-.04
-.03
.06
.28
.19
.03
.74
.03
.03
.63
.02
-.12
.17
.01
.11
.53
.01
13. A leader never shows weakness. 14. Leadership is about taking charge of a situation. 15. People can be taught to be more effective leaders. 16. Being a leader requires using both rewards and punishment. 17. Skills and abilities for leadership can be developed. 18. Leadership is a property of the group, not the individual. 19. Leadership is about maintaining control.
41 20. Leadership involves a group collectively making decisions. 21. Getting “buy-in” from followers is an important leadership task. 22. Effective leaders influence followers more than followers influence them.
.66
.06
.16
-.16
.16
.20
.23
-.08
-.09
-.02
.32
.10
23. Leaders, as people, are fundamentally different from non-leaders.
-.11
-.20
.52
.18
24. Leadership can be shared.
.39
.30
-.07
-.07
25. Leadership is the responsibility of everybody in a group.
.68
.01
-.09
.11
26. When a group performs well, it is largely because of the leader.
-.05
-.07
.39
.17
27. Leadership is about using influence.
-.08
-.00
.31
.16
.03
-.05
.16
.60
.13
.17
-.00
-.21
.08
.63
.05
-.13
.76
.06
.00
-.02
-.11
-.29
.51
.28
.22
.09
.09
.01
.08
-.64
.20
.17
.64
-.05
-.04
-.08
28. One's formal position determines whether they are a leader. 29. Leadership is a two-way street between leaders and followers. 30. Leaders can acquire skills to make them more effective. 31. Together, group members create leadership. 32. You're either a leader, or you're not. 33. Leaders allow followers to influence them. 34. You can't teach leadership. 35. Leadership is about the group, rather than a single leader.
36. Leadership is not possessed by any -.06 -.20 .49 one individual. 37. If you supervise others, you are a leader. .10 -.07 .08 Note. Bold factor loadings indicate a clear loading on that factor. N = 439.
.15 .46
Similar to Chan and Drasgow (2001), before subjecting the Orientation Toward Leadership measure to confirmatory factor analysis in the second (medical center) sample, a principal axis factor analysis (with varimax rotation) was performed on the medical center data. Eleven factors had eigenvalues over 1, although the scree plot suggested that three factors be retained (see Figure 6). These three factors corresponded to the three conceptual
42 leadership orientation dimensions and were the same factors as found in the student sample. Table 4 shows factor loadings for all items. Because of the similarity in items between the third and fourth factors in the student sample, and because items from both factors formed the third factor in the medical center sample, the third and fourth factor in the student sample were treated as the conceptual equivalent of a single factor. Items that loaded clearly and highly on the same factor in both samples were retained, leaving sixteen items (eight Shared Leadership, four Influence/Development, and four Dominance).
7 6
Eigenvalue
5 4 3 2 1 0 1
2
3
4
5
6
7
8
9
10
11
Factor
Figure 6. Scree plot of eigenvalues by Orientation Toward Leadership factor: Study 2.
43 Table 4 Results from Principal Axis Factoring of Orientation Toward Leadership Items: Study 2 Item 1. Telling other people what to do is a big part of leadership.
Shared -.15
Factor Loadings Development Dominance .02 .57
2. Group members can work together in performing leadership tasks.
.18
.45
-.11
3. Leaders order other people around. 4. Leadership involves persuading other people. 5. Effective leaders seek different perspectives. 6. Individual people do not possess leadership -- it is a property of the group. 7. Leaders should not let others openly question them. 8. Leadership happens when people collaborate.
-.16 -.10 .14 .45
.01 -.07 .20 -.01
.71 .10 -.04 .07
.08
-.09
.16
.67
.12
-.11
9. The leader should be the most influential person in a group. 10. Leadership is a process. 11. Leadership and power are pretty much the same thing. 12. All group members should participate in leadership. 13. A leader never shows weakness. 14. Leadership is about taking charge of a situation. 15. People can be taught to be more effective leaders. 16. Being a leader requires using both rewards and punishment. 17. Skills and abilities for leadership can be developed. 18. Leadership is a property of the group, not the individual.
-.10
.06
.48
.09 -.04
.32 -.22
.00 .75
.21
.11
-.18
.12 .05 .09 -.24
-.12 .02 .82 .13
.13 .18 -.09 .11
.12
.80
.01
.75
.06
-.07
19. Leadership is about maintaining control.
.07
-.02
.24
20. Leadership involves a group collectively making decisions. 21. Getting “buy-in” from followers is an important leadership task.
.75
.23
-.05
-.09
.13
-.00
-.15
.12
.21
22. Effective leaders influence followers more than followers influence them.
44 cont. Item
Shared
Factor Loadings Development Dominance
24. Leadership can be shared.
.20
.51
-.08
25. Leadership is the responsibility of everybody in a group.
.41
.18
-.01
26. When a group performs well, it is largely because of the leader. 27. Leadership is about using influence.
-.10
.05
.40
-.17
.13
.17
28. One's formal position determines whether they are a leader. 29. Leadership is a two-way street between leaders and followers.
.04
-.26
.57
.20
.28
-.06
30. Leaders can acquire skills to make them more effective. 31. Together, group members create leadership. 32. You're either a leader, or you're not.
.04
.73
-.04
.50 -.03
.25 -.16
-.02 .20
33. Leaders allow followers to influence them. 34. You can't teach leadership.
.14 -.07
.02 -.47
-.00 .11
35. Leadership is about the group, rather than a single .03 -.05 .83 leader. 36. Leadership is not possessed by any one .05 -.05 .41 individual. 37. If you supervise others, you are a leader. .09 .07 .41 Note. Bold factor loadings indicate a clear loading on that factor. Shaded items indicate that a particular item also loaded highly on the same factor in Study 1. N = 224.
In order to assess the consistency of the relationships between the resulting factors in student and medical center samples, correlations among factors were calculated for both samples (Table 5). Correlations were then transformed to z-scores using Fisher’s r to z′ transformation. The resulting z′ scores were then converted to a standard normal curve deviate (z) using the formula: z=
z′ Study1 – z′ Study 2 square root (1/[n1-3]+ 1/[n2-3])
This produced a standard normal curve deviate z value for each correlation comparison. The correlation between Shared and Dominance factors was not significantly different between Studies 1 and 2, z = -.49, p > .05. The correlation between Shared and Development factors
45 was not significantly different between Studies 1 and 2, z = -1.77, p > .05. Finally, the correlation between Development and Dominance factors was not significantly different between studies, z = 1.64, p > .05.
Table 5 Correlations Between Resultant Orientation Toward Leadership Factors: Study 1 and Study 2 Study 1 Shared Dominance Shared Dominance Development
-.10 .08
-.08
Study 2 Shared Dominance -.06 .22
-.21
Confirmatory factor analysis. The measurement model for the Orientation Toward Leadership measure was then assessed for fit using confirmatory factor analysis in LISREL 8 (Jöreskog & Sörbom, 1993). Missing data was imputed using multiple imputation and the EM algorithm. A series of three alternative nested models were assessed: a single-factor model (representing an overall world-view or method factor), a three-factor orthogonal model, and a three-correlated factors model. Although one generally expects factors to be at least slightly correlated with each other, it is plausible that they might not be correlated, or might be negatively correlated. For example, the Dominance factor may be negatively correlated with the Shared Leadership factor, and possibly the Influence factor. In order to better understand the relationship between these factors, these two different three-factor models were assessed. Results of the confirmatory factor analysis are presented in Table 6. Fit of the alternative models was assessed using both chi-square and practical fit indices, which are less sensitive to sample size. When evaluating a model, use of multiple indices is valuable because of the different assumptions of each of the indices (e.g., Chan & Schmitt, 1997). Practical fit indices include the root mean square error of approximation (RMSEA; Steiger, 1990), non-normed fit index (NNFI; Bentler & Bonett, 1980), comparative fit index (CFI; Bentler, 1990), and Jöreskog & Sörbom’s (1989) goodness of fit index (GFI). Nested models can be compared using the chi-square statistic. Models 2 and 3 are not nested within Model 1, and therefore Model 1 cannot be compared to Models 2 and 3.
46 Model 3 is nested within Model 2, allowing these models to be compared. An examination of the chi-square statistics of the two models (Table 7) shows that Model 3 fits the data significantly better than Model 2 (∆χ2 (3, N=231) = 28.48, p < .05). Model 3 specified that three orientation Orientation Toward Leadership factors (and their interrelationships) be estimated. Generally, RMSEA values less than .08 indicate good fit (Byrne, 1998). The RMSEA of .08 in Model 3 indicates good fit, although the RMSEA for the other two models (1 and 2) do not. Values above .90 for the NNFI, CFI, and GFI indicate good fit (Chan & Schmitt, 1997). The only model to exhibit good fit on both the NNFI and CFI was Model 3. None of the models exhibited satisfactory GFI values above .90; however Model 3 was the closest to this suggested cutoff. Taken together, Model 3 provided the best comparative fit to the data among the three models, and also provided satisfactory overall fit by measure of non-comparative fit indices. Figure 7 depicts the factors and their correlations based on results from Model 3.
47 Table 6 Confirmatory Factor Loadings for Leadership World-View Items Factor Loading Shared Development Dominance 18. Leadership is a property of the group, not the individual. 6. Individual people do not possess leadership -- it is a property of the group. 8. Leadership happens when people collaborate. 20. Leadership involves a group collectively making decisions. 25. Leadership is the responsibility of everybody in a group.
1.00 .50 .72 .83 .72
31. Together, group members create leadership.
.67
35. Leadership is about the group, rather than a single leader.
.89
36. Leadership is not possessed by any one individual.
.61
17. Skills and abilities for leadership can be developed.
1.00
15. People can be taught to be more effective leaders.
.99
30. Leaders can acquire skills to make them more effective.
.85
34. You can't teach leadership. 11. Leadership and power are pretty much the same thing. 3. Leaders order other people around. 28. One's formal position determines whether they are a leader. 37. If you supervise others, you are a leader. N = 224.
.40 1.00 .83 .77 .57
48 Table 7 Fit Statistics for Confirmatory Factor Analyses Comparing the Fit of Single-Factor Versus Two Alternative Three-Factor Models Model
Chi-square
df
RMSEA
NNFI
CFI
GFI
1. Single- factor model
935.35
104
0.19
0.62
0.67
0.66
2. Three-factor orthogonal model
294.54
104
0.09
0.89
0.90
0.86
3. Three-correlated factors model*
266.06
101
0.08
0.90
0.92
0.87
* Best-fit model. N = 231. RMSEA = root mean square error of approximation; NNFI = non-normed fit index; CFI = comparative fit index; GFI = goodness of fit index.
Dominance
-.20 Influence (Development)
-.21
.23 Shared Leadership
Figure 7. LISREL 8 correlations between factors for three-correlated factors model.
49 Assessing the measurement model. Once the measurement model has been established, Fornell and Larcker (1981) advocate testing the composite reliability and average variance extracted from the latent factors. The composite reliability calculation is analogous to coefficient alpha, but more appropriate in examining the reliability of items composing latent factors than coefficient alpha. Composite reliability is calculated using the formula: (Σ λ)2 (Σ λ) + Σ Var (ε) 2
,
where λ2 represents the standardized squared factor loadings and Var (ε) represents the variance due to measurement error. The reliabilities of the Shared Leadership, Development/Influence, and Dominance factors were .97, .80, and .93 respectively. These values indicate acceptable convergent validity of the items comprising each scale. The average variance extracted is a measure of the amount of variance captured by a construct in relation to variance associated with random measurement error. As such, it provides an assessment of construct validity. Variance extracted can be estimated using the formula: Σ (λ2) Σ (λ2)+ Σ Var (ε)
.
Values greater than .50 indicate that variance captured by the construct is greater than variance captured by measurement error and indicate convergent validity (as do high factor loadings on the intended construct in the confirmatory factor analysis). The average variance extracted for the three factors of Shared, Development/Influence, and Dominance Orientation Toward Leadership was .82, .50, and .76 respectively. This means that 82%, 50%, and 76% of the variance in the items comprising each construct are associated with the construct, and 18%, 50% and 24% (respectively) of the total variance in items from each construct is due to random measurement error. These values are at or above the cutoff of .50 recommended by Fornell and Larcker (1981) and help to provide evidence of the construct validity of the scales.
50 CHAPTER 6 Study 1: Method, Results, and Discussion Method Participants A total of 454 undergraduate students attending the Pennsylvania State University participated in Study 1. Fifty percent (228) were female. Participants were drawn from 5 separate courses (3 upper level psychology courses, one introductory psychology course, and one senior management course) and received course credit for their participation. Fifty-one percent (232) of the participants were from the Introductory Psychology Subject Pool, and 49% (222) from the other four participating courses. Data on participant age was not collected. Participation was voluntary, and the final list of participants was sent to each instructor by the researcher at the end of the semester, just before final grades were calculated. Procedure Students enrolled in the introductory psychology course were given the opportunity to sign-up for participation through the Psychology 002 Subject Pool. Students signed up for a specific available date and time, and, at the specific time, were sent an email containing an html (web) based survey (see Appendix E). Participants were instructed to complete the survey online within 24 hours. Approximately 93% of surveys sent were returned within 24 hours. After approximately 36 hours, an email was sent to students who signed up but failed to participate. After approximately 3 additional days, a final email was sent to students stating that failure to participate within the next 24 hours would result in a “no-show” whereby they would be penalized by receiving negative credit (in accordance with the Subject Pool requirements set forth by the Department of Psychology). Only 7 Psychology 002 students received a “no-show”. The sign-up procedure varied slightly for student participants in the Management and three other Psychology courses. Students were sent an email from their course instructor inviting them to participate in the research in exchange for extra credit. Students were given between 10 days and 3 weeks to participate. Before completing the survey, participants read an informed consent document, and “signed” by typing their name and date into response fields. Providing their name also allowed the researcher to ensure that credit was received for research participation. No other identifiers were collected. Instructions for completing the survey were provided as necessary, and participants were instructed to take a break at the half-way point in an attempt to mitigate fatigue effects.
51 Measures The full-length version of the measures used in Study 1 can be found in Appendix E. A description of the measures is presented below. Orientation Toward Leadership and Leadership Self-Identity. A Leadership SelfIdentity measure and Orientation Toward Leadership measure were developed for this study. Preliminary development and refinement of these measures are discussed in detail in Chapter 5. Reliability estimates of the dimensions of both measures were satisfactory. The Descriptiveness, Certainty, and Importance dimensions of Leadership Self-Identity had reliability values (coefficient alpha) of .92, .88, and .83 respectively. The Core Leadership Self-Identity score reliability was .93, and the reliability of the Overall Self-Identity (mean of all three dimensions) measure was .93. The coefficient alpha reliabilities of the three Orientation Toward Leadership scales -- Dominance, Development/Influence, and Shared – were .69, .77, and .84 respectively. Leadership experience. Previous leadership experience was conceptualized using Tesluk and Jacobs’s (1998) integrated and multidimensional model of general work experience, which delineates experience of three broad types: quantity, quality, and density/intensity. Although it is not known the extent to which measures of each of these components is differentially related to Leadership Self-Identity, the three types of experience are conceptually distinct. Indeed, all work and leadership experiences are not equal (McCauley, Ruderman, Ohlott, & Morrow, 1994; Quiñones, Ford, & Teachout, 1995). For example, an individual may spend twenty years in a leadership position, but the quality and intensity of his/her experiences may have been low. Thus, a measure of tenure may indicate a significant amount of experience, but the actual developmental impact of that experience may be low. Another individual may have only been in a leadership positions for a short duration but may have had high quality and intense and challenging experiences, providing him/her with a more developmental experience than the longer-year-tenured individual. In the leadership domain, different dimensions of experience have also been discussed, focusing largely on the Assessment, Challenge, and Support model of developmental experiences (Van Velsor & McCauley, 2004). This model holds that the highest quality experiences (i.e., those experiences that are the most developmental) have components of assessment, challenge, and support embedded in them. In the Tesluk and Jacobs (1998) framework, assessment and support could be generally classified as “quality” measures, and challenge aligns with the “density/intensity” component of experience. Combining both Tesluk and Jacobs’ (1998) model of experience and the Assessment, Challenge, and Support model of leadership experience, a multi-dimensional measure was created to assess the multiple domains of leadership experience. These dimensions were: (a) quantity, (b) quality – assessment, (c) quality – support, and (d) challenge/intensity. In terms of quantity of experience, relevant experience was believed to include the propensity to (a) be chosen as a leader, (b) be involved in leadership, (c) take on leadership responsibilities, and (d) be looked to by others for leadership. Quantity measures that assess
52 years of experience or number of leadership positions held were not chosen because of perceived difficulty in comparing different items within the quantity domain. The assessment dimension of experience quality is conceptually focused on ascertaining the current state of the individual’s leadership, and also where that individual should be or would like to be (Van Velsor & McCauley, 2004). Sources of assessment includes self-assessment, assessment from others, and feedback from the task itself. The second dimension of experience quality, support, concerns the environment within which development might occur. Support may come from a variety of sources (e.g., peers, family, subordinates, supervisors, organizational systems and cultures), but regardless of the source, support encourages and provides efficacy for development to occur. If a leadership experience takes place in an unsupportive environment – where change is not encouraged and carefully supported – the experience may not lead to development and may additionally not foster a leadership self-identity. Challenge and intensity of leadership experiences is presumed to be important because developmental leadership experiences are presumed to involve being forced out of one’s areas of comfort, engaging in new and challenging tasks, and being stretched to one’s limits (Van Velsor & McCauley, 2004). Using the conceptual descriptions of the different types of experience (quantity, quality [assessment and support], and challenge/intensity), items were written to capture each of these experience constructs. Results of principal axis factoring (with varimax rotation) suggest that the items loaded onto four distinct factors corresponding to the hypothesized dimensions of experience (see Appendix F for factor analysis results). Items that failed to load clearly (a single factor loading above .4 and no cross-loadings) were omitted from further analyses. The resulting measure was composed of 19 items: 4 quantity items, 5 feedback items, 6 support items, and 4 challenge/intensity items. Coefficient alphas for the four dimensions are: .88, .86, .89, and .80 respectively. Core self-evaluation. The Core Self-Evaluation Scale, developed by Judge, Erez, Bono and Thoresen (2003), is composed of thirteen items that assess individuals’ global, stable, evaluation of themselves. Conceptually, the core self-evaluation construct is a broadbandwidth construct representing the intersection among four widely-studied psychological traits: self-esteem, generalized self-efficacy, locus of control, and emotional stability (Judge, Locke, & Durham, 1997). Individuals with high core self-evaluations see themselves positively as agentic individuals, and are generally free from anxiety (Hiller & Hambrick, 2005). Responses on the scale range from strongly disagree (1) to strongly agree (5). Internal consistency reliability (α) for this scale was .83. Sample items are “When I try I generally succeed”, and “Overall, I am satisfied with myself.” Motivation to lead. Motivation to Lead (MTL) was measured using Chan and Drasgow’s (2001) 3-factor measure, with a total of 27 items (9 per factor). Affective-identity MTL is defined by liking to lead others. Noncalculative MTL assesses the calculative beliefs about the costs of taking on leadership, whereas less calculative individuals take on leadership roles because they do not calculate the burden of leadership. Social-normative
53 MTL taps the sense of duty and responsibility about leading that an individual possesses. Coefficient alpha reliabilities for the three factors were .90 for affective-identity MTL, .84 for social-normative MTL, and .70 for noncalculative MTL. Participants responded on a five-point Likert-type scale ranging from strongly disagree (1) to strongly agree (5). Sample items for the affective-identity, social-normative, and noncalculative factors are, respectively: “I am the type of person who likes to be in charge of others”, “I would agree to lead others even if there are no special rewards or benefits with that role”, and “I feel that I have a duty to lead others if I am asked.” Self-monitoring. Lennox and Wolfe’s (1984) 13-item measure of self-monitoring was used to assess the extent to which an individual observes, regulates, and modifies the “self” that is presented to others. Participants responded on a five-point Likert-type scale, ranging from strongly disagree (1) to strongly agree (5), to items such as, “In social situations, I have the ability to alter my behavior if I feel that something else is called for.” In the present study, coefficient alpha was .87. Leadership self-schema (trait method). Leadership self-schema was assessed using a protocol similar to that used by Engle and Lord (1997). In essence, this general approach determines traits that are characteristic of leaders, and then asks participants to rate how characteristic each trait is of themselves. In the present study, participants were asked to assess 21 traits according to how characteristic of a leader the traits were (using a list of implicit leadership traits developed by Epitropaki & Martin, 2004). In a later section of the survey (in order to attempt to decrease priming and response bias), participants were asked how characteristic each of the 21 traits was of themselves. Examples of traits include those that are generally prototypical of leaders (understanding, knowledgeable), and traits that are antiprototypical (domineering). Participants responded on a five-point Likert-type scale, ranging from not at all characteristic (1) to completely characteristic (5). In order to calculate a leadership self-schema score, mean scores for each of the 21 traits were calculated across all participants to determine which traits and characteristics were widely seen by the sample as being characteristic of a leader. Ten of the 21 traits were rated consistently as being characteristic of a leader, with an overall item mean of 4.0 or higher on a 5 point scale (Engle & Lord, 1997). These traits were: understanding, helpful, dynamic, intelligent, knowledgeable, educated, dedicated, motivated, hard-working, and energetic (see Appendix C for item-level means)9. These 10 widely endorsed leadership traits were then used to analyze participants’ self-assessments on these same dimensions. According to Engle and Lord (1997), an overall mean self-rating for each individual on these widely endorsed leadership dimensions is an accurate reflection of one’s leadership self-schema. Coefficient alpha reliability for the 10item leadership self-schema measure was .83.
9
Antiprototypical leader traits such as pushy, loud, and manipulative were also assessed for inclusion as a negatively loading implicit leadership theory factor. None of the items, however, displayed sufficiently low means (below 2.0 on a 5 point scale) to be considered as broadly antiprototypical.
54 Individualism and collectivism (horizontal and vertical). Two unique measures of individualism (horizontal and vertical) and two measures of collectivism (horizontal and vertical) were assessed using 27 items that displayed adequate factor loadings in a study by Singelis, Triandis, Bhawuk, and Gelfand (1995). Participants were asked to assess their level of agreement with each statement on a 7-point scale, from strongly disagree (1) to strongly agree (7). The measure of horizontal individualism consisted of 5 items, such as “I often do my own thing” (α = .68). Eight items made up the vertical individualism measure, with items such as “When another person does better than I do, I get tense and aroused” (α = .79). Horizontal collectivism was assessed with 8 items, including “The well-being of my coworkers is important to me” (α = .74). Vertical collectivism was preliminarily assessed using 6 items, such as “I would do what would please my family, even if I detested that activity”. After examining the poor reliability of the six-item vertical collectivism scale (α = .50), two items were selected for removal based on principal axis factoring (with varimax rotation) of the six items. The removal of these two items (“I hate to disagree with others in my group”, and, “Children should feel honored when their parents receive an award”) increased the reliability of the scale (α) to .63. Self confidence in leadership (leadership self-efficacy). Self-confidence in leadership was assessed using 7 items, based on measures developed by Feasel (1995) and Chan and Drasgow (2001). Items were assessed on a 1-7 scale, with confidence anchors ranging from “0% confidence” (1) to “100% confidence” (7). A sample item is “I am confident that I can lead others effectively”. Principal axis factoring (with varimax rotation) found that all seven items loaded highly on one factor (interpreted as a self-confidence in leadership factor), which accounted for 73% of the variance in items. Coefficient alpha reliability for this scale was .94. The full-version of the scale can be found in Appendix E.
Results Means, standard deviations, and correlations among study variables are presented in Table 8. Hypotheses 1 though 5 predicted relationships between leadership self-identity and various individual difference variables, including previous leadership experience, personality variables, and motivation. Each hypothesis was examined using data from Study 1. Hypothesis 1 predicted that previous leadership experience is related to seeing oneself as a leader (leadership self-identity). As Table 8 shows, all three subcomponents of leadership self-identity (descriptiveness, certainty, and importance) were significantly related to all four measures of previous leadership experience (quantity, assessment, support, and challenge/intensity), thus Hypothesis 1 was fully supported. Quantity of previous leadership experience (measured by four items assessing how often the participant engaged in leadership) accounted for 56% of the variance in the Descriptiveness component of Leadership Self-Identity, r(455) = .75, p < .01. Bivariate correlations among Leadership Self-Identity dimensions and other measures of previous leadership experience (feedback and
55 mentoring, developmental support, and challenge/intensity of leadership experiences) were also significant, ranging from .28 to .47 (all p < .01). As a further examination of Hypothesis 1, a multiple regression was performed. Regressing Leadership Self-Identity (descriptiveness) on all four experience variables simultaneously revealed that quantity of leadership experience, feedback and assessment during previous leadership experiences, and support and development during previous experiences were significant contributors to Leadership Self-Identity (Table 9). Challenge/intensity of previous experience was not a significant predictor when considered simultaneously with other experience measures. Hypothesis 2 predicted that an individual’s core self-evaluation is positively related to leadership self-identity. In other words, individuals with more globally positive evaluations of themselves are more likely to see themselves as leaders. As hypothesized, correlations from Table 8 show that an individual’s core self-evaluation (CSE) is significantly related to all three subcomponents of Leadership Self-Identity (descriptiveness, certainty, and importance). Thus, Hypothesis 2 is fully supported. Hypothesis 3 argued that differential relationships might exist between leadership self-identity and the three dimensions of the motivation to lead (MTL) construct. Affectiveidentity MTL, which assesses the extent to which an individual likes to lead others, was hypothesized to be positively related to seeing oneself as a leader (leadership self-identity; Hypothesis 3a). Similarly, leading out of a sense of duty (social-normative MTL) was hypothesized to be positively related to seeing oneself as a leader (Hypothesis 3b). Bivariate relationships between affective-identity MTL and the three Leadership SelfIdentity dimensions were particularly strong, ranging from .49 to .73 (all p < .01). Bivariate relationships between social-normative MTL and the dimensions of Leadership Self-Identity (descriptiveness, certainty, and importance) were also significant, ranging from .22 to .41 (all p < .01). Thus, Hypotheses 3a and 3b are supported: liking to lead others and leading out of a sense of duty is significantly related to Leadership Self-Identity. Hypothesis 3c posited a negative relationship between noncalculative MTL and leadership self-identity. Bivariate correlations, however, failed to support Hypothesis 3c. In fact, noncalculative MTL was significantly related to Leadership Self-Identity, with correlations between .09 and .20 (Table 8). According to the bivariate relationships, individuals who were noncalculative about the costs of leadership were more likely to see themselves as leaders (Leadership Self-Identity).
Table 8 Correlations Among Variables for Student Sample Scale
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
LSI (descriptiveness) LSI (certainty) LSI (importance) LSI first factor LSI grand mean Experience: Quantity Experience: Feedback Experience: Support Experience: Challenge/Intensity Core self-evaluation Affective-identity MTL Non-calculative MTL Social-normative MTL Self-monitoring Leadership self-schema OTL: Shared OTL: Dom. OTL: Develop. Horizontal individualism Vertical individualism Horizontal collectivism Vertical collectivism Self-confidence in leadership
M 4.50 4.29 3.61 3.75 3.93 3.56 5.08 5.47 5.11 4.87 4.76 5.19 4.75 5.22 4.10 4.33 3.77 5.09 5.36 4.34 5.34 4.01 5.73
SD
.98 .91 1.08 .87 .77 .71 .90 .95 1.02 .78 1.06 .89 .68 .73 .47 .91 .85 .58 .75 .92 .67 .98 .88
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
.92 .60 .70 .93 .91 .75 .46 .44 .47 .25 .73 .18 .36 .32 .47 -.06 -.02 .18 .29 .32 .14 .10 .62
.88 .46 .58 .80 .53 .30 .27 .30 .22 .49 .20 .22 .23 .30 -.08 -.09 .16 .24 .14 .13 .06 .38
.83 .91 .84 .52 .29 .25 .37 .09 .54 .09 .41 .25 .33 -.07 .07 .15 .19 .35 .13 .18 .42
.93 .95 .70 .41 .39 .46 .18 .69 .15 .42 .31 .43 -.06 .02 .18 .26 .36 .15 .12 .57
.93 .71 .42 .38 .45 .22 .69 .19 .39 .32 .43 -.08 -.02 .19 .29 .32 .16 .11 .56
.88 .41 .38 .54 .24 .70 .25 .28 .31 .42 -.04 -.07 .13 .31 .24 .14 .05 .58
.86 .65 .46 .33 .31 .25 .24 .36 .45 .19 -.02 .30 .28 .11 .27 .05 .44
.89 .44 .27 .34 .27 .27 .29 .41 .10 -.03 .37 .22 .06 .28 .09 .38
.80 .05 .42 .21 .23 .29 .36 .11 .00 .20 .20 .21 .23 .11 .37
.83 .31 .29 .17 .24 .42 -.07 -.14 .18 .26 .05 .11 -.09 .43
.90 .38 .37 .36 .36 -.13 -.05 .09 .32 .31 .08 -.02 .61
.84 .21 .24 .35 -.06 -.33 .12 .21 -.20 .35 -.08 .32
.70 .29 .27 -.12 .11 .14 .22 .29 .22 .17 .37
.87 .31 .01 .05 .24 .34 .17 .24 .04 .43
.83 .09 -.10 .28 .35 .07 .40 .07 .48
.84 -.10 .08 -.04 -.17 .20 .13 -.03
.69 -.08 .03 .25 -.12 .05 -.06
.77 .20 .03 .31 .08 .16
.67 .19 .27 .01 .37
Correlations at or above .08, p < .10. Correlations at or above .09, p < .05. Correlations at or above .12, p < .01. LSI = Leadership Self-Identity, MTL = Motivation To Lead, OTL = Orientation Toward Leadership .
20
21
22 23
.79 -.13 .74 .20 .22 .63 .20 .20 .04 .94
N = 446-454. 56
57 Table 9 Multiple Regression Results of Previous Leadership Experience Dimensions Predicting Leadership Self-Identity Leadership Experience Variable
Standardized Betas
Leadership Frequency Feedback & Assessment Support & Development Challenge & Intensity
.655** .105** .117** .017
R2 Adjusted R2 F a df = 4, 442 * p < .05.
.60 .60 164.83** ** p < .01
High status is associated with leadership (Hogan & Hogan, 2002), and because high self-monitors are likely motivated by the appearance of being “high-status” (Gangestad & Snyder, 2000), it was hypothesized (Hypothesis 4a) that high self-monitors would see leadership as being descriptive of themselves (i.e., high Leadership Self-Identity descriptiveness). The data support Hypothesis 4a, r(450) = .32, p < .01: self-monitoring was positively related to the descriptiveness dimension of Leadership Self-Identity. Hypotheses 4b and 4c posited inverse relationships between self-monitoring and both leadership self-certainty and leadership self-identity importance because (a) high selfmonitors may have more difficulty thinking about themselves, and (b) self-knowledge and attitudes of high self-monitors are less accessible than those of low self-monitors (DeBono, Green, Shair, & Benson, 1995; Kardes, Sanbonmatsu, Voss, & Fazio, 1986; Snyder & Cantor, 1980). Results, however, failed to support Hypothesis 4b and Hypothesis 4c. Bivariate correlations between self-monitoring and both Leadership Self-Certainty and Leadership Self-Importance were significant and positive, rather than negative, r(450) = .23, p < .01 and r(448) = .25, p < .01. It was hypothesized that leadership self-identity would be positively related to assessments of the self on prototypical leader traits. Results support the notion that these concepts are related; leadership self-schema was positively related to all three dimensions of Leadership Self-Identity (Table 8), with correlations between .30 (certainty dimension of LSI; p < .01) and .47 (descriptiveness dimension of LSI; p < .01). Thus, hypothesis 5, that leadership self-schema and Leadership Self-Identity are positively related, was supported.
58 Testing Antecedents/Correlates of Leadership Self-Identity in a Single Model After developing a better understanding of the bivariate relationships between Leadership Self-Identity and the five broad conceptual correlates/antecedents (previous experience, core self-evaluations, motivation to lead, self-monitoring, and leadership selfschema), the relative contribution of the set of predictors in explaining the three Leadership Self-Identity dimensions was sought through three simultaneous multiple regressions (Table 10). Two predictors (previous quantity of leadership experience, and affective-identity motivation to lead) were consistent predictors of all three dimensions of Leadership SelfIdentity. On the other hand, both self-monitoring and the challenge/intensity component of previous leadership experience were not significant predictors of any of the dimensions of Leadership Self-Identity. Other predictors varied according to the self-identity dimension in question. In terms of overall variance explained, 70% of the variance in the descriptiveness dimension of Leadership Self-Identity was explained by the set of all predictors. The set of predictors as a whole explained 30% of the variance in identity certainty and 42% of the variance in identity importance.
Table 10 Multiple Regression Results Examining Antecedents to Leadership Self-Identity: Descriptiveness, Certainty, and Importance a Predictor Leadership Experience: Quantity Leadership Experience: Feedback Leadership Experience: Development and Support Leadership Experience: Challenge and Intensity Core Self-Evaluation MTL: Affective-Identity MTL: Noncalculative MTL: Social-Normative Self-Monitoring Leadership Self-Schema
Leadership Self-Identity Descriptiveness Certainty Importance .37** .12** .07*
.31** .09 -.01
.20** .09† -.05
-.02
-.02
.03
-.05 .43** -.17** .07* -.02 .14**
.03 .23** .00 .02 .01 .04
-.15** .35** -.17** .24** .00 .13**
.32 .30 20.07**
.43 .42 33.22**
.71 R2 Adjusted R2 .70 b F 105.08** a standardized regression coefficients are shown. b df = 10, 436. † p < .10. * p < .05. ** p < .01
59 Orientation Toward Leadership A number of relationships between Orientation Toward Leadership dimensions (Dominance, Development/Influence, and Shared) and the four dimensions of individualism and collectivism were also predicted (Hypothesis 6a-h). Vertical individualism was positively related to Dominance, r(449) = .25, p < .01, and negatively related to shared leadership orientation, r( 449) = -.17, p < .01, supporting Hypothesis 6a and 6b. Hypothesis 6c was not supported: horizontal individualism was not related to shared leadership orientation, r(449) = -.04, p > .10. Hypothesis 6d was supported: horizontal individualism was positively related to development orientation toward leadership, r(448) = .20, p < .01. Relationships between horizontal collectivism and the shared dimension of leadership orientation were predicted in Hypothesis 6e. The correlation was significant and positive, r(448) = .20, p < .01, thus supporting Hypothesis 6e. Horizontal collectivism was negatively related to a dominance leadership orientation, r(448) = -.12, p < .01, supporting Hypothesis 6f. Finally, vertical collectivism was posited to be positively related to both shared leadership orientation (6g) and Dominance leadership orientation (6h). Vertical collectivism was positively related to shared orientation, r(447) = .13, p < .01, supporting Hypothesis 6g. Hypothesis 6h was not supported; vertical collectivism was not related to dominance leadership orientation. Taken together, six of the eight hypothesized relationships between individualism/ collectivism and the three leadership orientations were significant in the direction predicted.
Discussion Study 1 was designed to understand individual difference correlates of leadership identity and views of leadership, and in doing so help provide convergent and discriminant validity evidence for both measures. Results help to paint a picture of the individual high/low on leadership self-identity and high/low along the three leadership orientation dimensions. Leadership Self-Identity Correlation and regression analyses showed that whether one sees the self as a leader (leadership self-identity) was related to almost all of the individual difference variables assessed, although in a few cases, results were in the direction opposite to that predicted. LSI
60 dimensions were most highly correlated with quantity of previous leadership experience and affective-identity motivation to lead (MTL). In other words, individuals who want to lead others and who have previous experience in leadership are likely to see themselves as leaders. Significant positive correlations were also observed with other experience variables, such as feedback about experience, support during experience and having challenging leadership experiences. Having a positive overall evaluation of the self (core self-evaluation) was also related to seeing the self as a leader, as was feeling obliged to lead (socialnormative MTL). In further understanding the Leadership Self-Identity measure developed in the present research, comparisons were made between Leadership Self-Identity and the previously existing measure of leader perceptions of the self (“Leadership self-schema”; Engle & Lord, 1997). The measure of leadership self-schema developed by Engle and Lord has been argued to be an indicator of the identity one has around leadership. A careful and conceptual analysis of their measurement technique suggests that it is a self-assessment of the traits associated with leaders; not conceptually equivalent to a directly-measured leadership identity. The Leadership Self-Identity measure developed in the present research was believed to be a more direct assessment of the self-identity one holds in the leadership domain by asking directly about leadership self views. Although conceptually similar and probably empirically related to Engle and Lord’s schema measure, the correlations were predicted to be moderate. If correlations between leadership self-schema and leadership self-identity dimensions (or aggregate scores) were substantially high, an argument might be made that these methods of measurement are practically equivalent. As predicted, leadership self-schema was positively related to leadership self-identity. Individuals who perceive themselves as holding prototypical leader traits are more likely to see themselves as leaders. However, at the same time these measures are related, results showed self-schema scores only account for between 9% and 22% of the variance in selfidentity scores (depending on the dimension of self-identity). This adds empirical evidence on top of the conceptual argument that these two measures are related but not equivocal and thus provides another piece of the evidence in establishing both convergent and discriminant validity of the Leadership Self-Identity measure. Several results in the student sample ran contrary to hypotheses. Self-monitoring was believed to have differential relationships with the different dimensions of leadership selfidentity. Results show, however, that self-monitoring was positively correlated with all three dimensions such that high self-monitors see leadership as being descriptive of themselves, they are certain of this identity, and further, that this identity holds higher self-importance than for low self-monitors. It has previously been found that high self-monitors have more difficulty thinking about themselves and that self-knowledge and attitudes of high selfmonitors are less accessible than those of low self-monitors (DeBono, Green, Shair, & Benson, 1995; Kardes, Sanbonmatsu, Voss, & Fazio, 1986; Snyder & Cantor, 1980), which led to the hypothesis of a negative relationship between self-monitoring and the certainty and importance dimensions of leadership self-identity. However, the appearance of being a leader, which is a “high status” position, may have overridden any negative influence on the certainty and importance dimensions of self-identity. Further, changing the presentation of
61 self in different situations (which is the essence of high self-monitoring) does not preclude a core self-identity that these individuals hold as important, are certain of, and able and willing to report in “safe” situations such as an anonymous survey. The other analyses that ran counter to the expected direction were the set of correlations between self-identity for leadership (all three dimensions) and noncalculative MTL. According to Chan and Drasgow (2001), noncalculative MTL presumes that people are more likely to lead if they do not calculate the costs and burdens that are associated with taking a leadership position. From the self-schema literature, it is known that individuals who are schematic in a given domain spend more time thinking about that domain (see Cross & Markus, 1994). As a result, individuals with a strong leadership self-identity were expected to spend time and resources thinking about leadership, including personal costs of leadership (calculative MTL). As a result, it was hypothesized that stronger self-identity around leadership would be related to lower noncalculative MTL (i.e., more calculative about leadership). The data, however, did not show this relationship; leadership self-identity was positively related to noncalculative MTL. In other words, having a leadership self-identity was inversely related to thinking about the costs of leadership. One possible explanation of this finding comes from a careful examination of the items composing this measure, which paints a slightly different picture of noncalculative MTL than that provided by Chan and Drasgow (2001). Eight of the nine noncalculative MTL items focus on the extent to which people possess a desire to lead only when there are clear benefits and/or minimal costs (e.g., “I would only agree to be a group leader if I know I can benefit from that role”, and “I am only interested to lead a group if there are clear advantages for me”). Thus, the emphasis of the items is on leadership as only an economictype transaction with selfish motives. Rather than noncalculative MTL centering on ignorance of leadership costs which are presumed to be part of leadership roles (Chan & Drasgow, 2001, p. 482), it is more appropriately described as the degree to which one is motivated to engage in leadership only when the transaction of costs and benefits has net positive results. An individual with a strong leadership self-identity sees the self as a leader and wants to lead (as evidenced by relationships between affective-identity MTL and leadership selfidentity). In fact, they may be partially motivated to lead because they want to express their identity as a leader. As a result, it could be that individuals with a strong leadership selfidentity do not focus on leadership as a transaction of costs and benefits. There may be benefits and costs to leading that are recognized by the individual high in leadership selfidentity, however these cost/benefit motives may be of secondary importance and/or more complex than captured in the noncalculative MTL measure. Before spending too much time in interpretation, however, it should be recognized that a different picture emerges about noncalculative MTL when examining all individual difference variables simultaneously in a regression equation. Controlling for all other variables (and thus, controlling for some of the shared variance between noncalculative MTL and other predictors), noncalculative MTL was negatively related to self-identity (descriptiveness and importance dimensions). This result, which conflicts with the bivariate
62 correlations, provides a picture of that part of noncalculative MTL which is unique because it is controlling for variance shared with other variables. The extent to which the unique component of noncalculative MTL tested in this regression equation is an accurate reflection of the true underlying construct is unknown, but it may be that the essence of noncalculative MTL is perhaps truly negatively related to leadership self-identity. The extent to which this assumption is accurate needs further investigation through continued construct validation; however the possibility is intriguing and further suggests that the measure of noncalculative MTL may need additional conceptual clarification and refinement. Orientation Toward Leadership Orientation toward leadership dimensions, which assess three separate fundamental beliefs about the nature of leadership, were predicted to be correlated with vertical and horizontal components of individualism and collectivism. However, rather than a uniform relationship between all Orientation Toward Leadership factors and each individualism/collectivism measure, some relationships were projected to be positive, while others were projected to be negative. Individuals high on vertical individualism are defined by an emphasis on the self, competitiveness, and acceptance of power and status differences (Singelis et al., 1995). Not surprisingly, these individuals were likely to endorse a notion of leadership as being about dominance (Hypothesis 6a) and were be less likely to endorse the possibility that leadership can be effectively shared among members of a group (Hypothesis 6b). Horizontal individualists, who similarly emphasize the self but are more focused on equality between people, did not endorse dominance (Hypothesis 6c) or shared approaches to leadership, but were more likely to hold a belief that leadership can be trained and developed (Hypothesis 6d). Individuals high in horizontal collectivism are defined by an emphasis on the group and equality between all group members. Not surprisingly, high horizontal collectivism was positively related to shared leadership orientation (6e) and negatively related to dominance orientation which emphasizes power and status differences between leaders and followers (6f). The finding that vertical collectivism was positively related to shared leadership orientation was hypothesized (6g), although results failed to support a relationship between vertical collectivism and dominance. Examined slightly differently, one additional pattern is worth mention. The belief that leadership can be taught (high development) was positively related to both horizontal individualism and horizontal collectivism. Although individualism and collectivism place differing value on the extent to which the self is seen as an individual or as part of a group, the horizontal components that both dimensions share are the rejection of status differences and the value placed on equality. But why should “equality” relate to a developmental view of leadership? One possibility is that differences in innate ability are not emphasized in a horizontal orientation, which allows for the possibility that leadership can be taught and
63 developed within all individuals. Further exploration of this finding may help to shed further light on this idea. Taken together, there was general support for the existence of relationships between beliefs about leadership and both the extent to which people support notions of equality vs. competitiveness, and how people construe themselves in relation to others (individualism and collectivism). Six of the eight hypothesized relationships were significant in the expected direction. Although results were supportive of hypotheses, only two of the correlations were .25 or higher. Thus, while providing evidence of convergent validity, there are other factors that are likely to be related to leadership orientation. A post-hoc examination of the correlations between other variables and leadership orientation reveals that a number (Table 8) of individual difference variables were significantly predictive. Notably, a dominance leadership orientation was inversely related to noncalculative motivation to lead. Individuals who see leadership as involving dominance are also likely to be more willing to lead only when there are clear benefits to leading. Developing leadership in high dominance individuals could involve an attempt to focus on the benefits of leadership (Day, Sin, & Chen, 2004) or might involve de-emphasizing the transaction of tangible costs and benefits altogether. Two other noteworthy post-hoc correlations were found between development leadership orientation and two leadership experience components (Table 8). Having both (a) supportive and developmental experiences and (b) receiving feedback and assessment about leadership experiences was significantly related to viewing the potential for leadership to be developed and trained. These correlations were higher than those found for both quantity of experience and challenge/intensity of experience. This provides additional evidence to the basic argument that all experiences are not equal (McCauley, Moxley, & Van Velsor, 1998). In addition, these findings suggest that seeing the potential for leadership to be something that can be effectively trained and developed may be particularly related to high quality experiences that include support, purposeful development, feedback, and assessment during and after leadership experiences. Although caution should be used in drawing causal conclusions because of the correlational nature of this study, behavioral experiences around leadership may alter the cognitions one holds about the source and malleability of leadership. Future research using time-lagged data might examine temporal dominance to better understand the relationship between leadership behaviors and cognitions. Relationship between Leadership Self-Identity and Orientation Toward Leadership In Study 1, relationships between the three leadership self-identity components and the three orientation toward leadership dimensions were weak. Of nine correlations, four were statistically significant (p < .05), but the strongest correlation between measures of selfidentity and leadership orientation accounted for only 3.2% of the variance, indicating that these variables may not be practically significant. These low correlations are noteworthy because they suggest that seeing the self as a leader (or not) is essentially unrelated to what are one’s cognitive views of the source and nature of leadership (orientation toward leadership). Having a highly self-important leadership identity does not mean that one is
64 necessarily highly dominant; nor does it suggest that a strong leadership self-identity prevents one from valuing a shared approach to leadership. While both may act in concert to predict relevant leadership behaviors, they are not codependent.
Brief Summary Orientation toward leadership and leadership self-identity were generally related to conceptual correlates and antecedents as predicted. Combined, the results help to provide a rudimentary understanding of the relationships between leadership orientation and self-identity with other variables, and in doing so, help to provide an understanding of the constructs themselves. Study 2 was designed to further examine these constructs and understand their importance by assessing outcomes of leadership self-identity and leadership orientation in a field setting.
65 CHAPTER 7 Study 2: Method, Results, and Discussion Method Participants The sample for Study 2 was composed of medical and nurse supervisors and their subordinates at the Penn State Milton S. Hershey Medical Center (PSHMC), a teaching hospital affiliated with The Pennsylvania State University. The organizational context was leadership development intense; a variety of programs for leadership development and an inhouse leadership development center existed, contributing to a culture that promoted leadership knowledge and leadership development. All participants were either clinicallybased medical faculty or nurses. Nurse supervisors held the position of either nurse manager or clinical head nurse. Supervisors, clinical division chiefs, and clinical department chairs comprised the group of clinical faculty supervisors. All clinical faculty supervisors and all nurse supervisors were sampled, and up to 10 subordinates per supervisor were sampled (mean = 5.72, range 1 - 10). Because supervisor participants had to provide ratings of each employee, it was determined that assessments of more than 10 subordinates would be burdensome. In cases in which more than 10 subordinates were overseen by a given supervisor, 10 employees were randomly selected by the HRIS manager. In cases where a supervisor supervised less than 10 employees, all of the subordinates were sampled. In total, surveys were mailed to 102 supervisors (48 clinical faculty supervisors and 54 nurse supervisors) and 583 subordinates. Six subordinate surveys were returned by the postal service as undeliverable. Three blank surveys were returned to the researchers by participants, and four participants noted that the supervisor listed was not their supervisor. No supervisor surveys were returned blank or undeliverable. At the request of the participating organization and to satisfy requirements of the Institutional Review Board at PSHMC, surveys were sent by mail to the home address of participants. Forty-four usable supervisor surveys were returned to the researcher, resulting in a supervisor response rate of 43.1%. These 44 supervisors rated a total of 245 employees. Usable subordinate surveys were returned by 187 employees, resulting in a response rate of 32.6% (187/573). For some analyses, matching supervisor and subordinate data were required. The total number of cases where both a supervisor and corresponding subordinate provided data was 163 (36 supervisors and 127 subordinates). Of the 44 supervisors providing data, 3 (7.3%) were between 31-40 years old, 17 (41.5%) were between 41 and 50 years old, 18 (43.9%) were between 51 and 60, and 3 were
66 over 60 years old. Average tenure in current position was 12.8 years. Two supervisors did not report age or tenure. Of the 187 subordinates who returned surveys, 19 (10.6%) were between 21 and 30 years old, 38 (21.2%) were between 31 and 40 years old, 58 (32.2%) were between 41 and 50 years old, 54 (30.0%) were between 51 and 60 years old, and 11 (6.1%) were over 60 years old. Average tenure in current position was 8.5 years. Seven subordinates did not report age or tenure. Sample background. The predominant responsibilities of the clinical medical faculty within PSHMC involve patient care (i.e., they do not spend a majority of their time on research and/or teaching as faculty in other areas of the university typically would). Research faculty members generally have less contact with their supervisors and are significantly autonomous in their work, and thus were not sampled in this study. Clinical faculty members are relatively autonomous, but have more consistent contact with their supervisors and were thus a more appropriate sample for this study. Based on discussions with members of the Human Resources Department, the demographic makeup of this sample was estimated to be similar to the national averages for medical college faculty: 72% male, 75% White (non-Hispanic), 12% Asian American, and 3% African American (American Association of Medical Colleges, 2003). The predominant responsibilities of nurses range widely by department, although patient care is a commonality to all nurse positions. Demographics of nurses were not assessed, though discussions with the Human Resources Department led to an estimation that the demographic makeup of this sample conformed to national norms Nationally, 94.1 percent of employed nurses are female and 10.2 percent have a master’s or doctorate in nursing (Department of Health and Human Services, 2001). Nearly 87% of nurses are white (non-Hispanic), 5% African American, 4% Asian, and 2% Hispanic. Selecting the appropriate supervisor-subordinate pairing. Surveys of supervisors and their direct reports needed to be compared in order to test some of the hypotheses, which required a careful determination of the appropriate supervisor-subordinate pairings. In most cases, the appropriate pairing was easily determined. In a number of cases where the appropriate pairing was not easily determined, decisions as to the most relevant pairing were made by a Human Resource Information Systems (HRIS) manager at PSHMC, in conjunction with the Chief of Human Resources. In the few cases where multiple supervisors existed, the supervisor with the most direct employee contact was chosen. Procedure Following Dillman’s (2000) research-based suggestion for achieving a higher response rate to mail surveys, participants received three contacts about the process, including: (a) presurvey contact, (b) survey, and (c) thank you and reminder. The presurvey contact (see Appendix G) was a brief letter informing participants about the study and informing them to expect to receive the actual survey in 2-4 days. To
67 increase perceived legitimacy of the study by participants (which, ultimately, should lead to a higher response rate), the pre-survey letter was sent under the auspices of the chair of the Department of Surgery at the Penn State Milton S. Hershey Medical Center and a professor of Industrial/Organizational Psychology at Penn State (University Park campus) who is involved in leadership development efforts at PSHMC. The actual survey packet, which was sent 3 days after the initial pre-survey letter, contained the survey (see Appendices H and I), a Summary Explanation of Research (see Appendix J), and a business reply envelope (postage paid by metered mail). The survey was kept to a minimum length to encourage participation; a particularly important consideration given the nature of the sample and the organizational climate. It was estimated that participation of subordinates required approximately 15 minutes of time. Participation of supervisors required approximately 10-20 minutes of time, depending on the number of subordinates (from 1 to 10) that were rated. Seven days after the surveys were mailed a follow-up letter was mailed to participants (see Appendix K). The purpose of this generic letter was to thank those who had completed the survey and to remind individuals who had not completed the survey that their participation was valued. In addition, contact information for the researchers was provided so that individuals who had misplaced their survey could receive a replacement. Six individuals requested a replacement survey. The letter also contained a survey completion deadline of 8 days later, which was given to convey the impression of importance and urgency. Many surveys were received after the stated deadline, and were included in the analyses. Pre-coding of surveys. While extensive care was taken to ensure participant confidentiality, anonymity of participant responses was not possible because of the need to match subordinate responses with their appropriate supervisor responses. Based on requirements from the Institutional Review Board at the Hershey Medical Center, a blind coding scheme was developed so that the researcher would not have any way to link a respondent’s name to a specific survey. After receiving the full list of all individuals who were selected as potential participants (previously determined by an HRIS manager at PSHMC, and including all supervisors and up to 10 randomly selected subordinates per supervisor), a set of random alphanumeric codes was created and paired with individual names. Each survey was coded with the random alphanumeric code for the participant, as well as the code of the corresponding supervisor or subordinate that each employee was to assess. On top of the alphanumeric code of each rated individual, a strip of paper was attached that identified the name of the corresponding supervisor or subordinate that each employee was to assess. This ensured that there was no confusion as to who the participant was assessing. Participants were instructed to remove the strip(s) of paper once the survey was completed so that only the alphanumeric code remained. Thus, surveys returned to the researcher contained only alphanumeric codes of raters and ratees. All coding was done in the executive office area of PSHMC, and was supervised by the Chief of Human Resources and the Director of PSHMC’s Leadership Institute. The master coding sheet, which contains both names of participants and matching codes, was retained by the Chief of Human Resources. A modified coding sheet was made available to the researcher, which contained
68 only an indication of code pairings so that supervisor and subordinate surveys could be matched appropriately while maintaining anonymity. Measures The full-length version of the measures used in Study 2 can be found in Appendices H (supervisor survey) and I (subordinate survey). Leadership self-identity and orientation toward leadership. A full description of the development and refinement of the Leadership Self-Identity measure and the Orientation Toward Leadership measure is discussed in detail in Chapter 5. Reliability of these scales was satisfactory. The descriptiveness, certainty, and importance dimension reliability (alpha) values were, respectively, .87, .95, and .93. The Self-Identity Factor score reliability was .91, and the reliability of the Overall Self-Identity score was .90. The coefficient alpha reliabilities of the three Orientation Toward Leadership scales – Dominance, Development/Influence, and Shared – were .70, .77, and .84, respectively. Leader-member exchange. Based on the results of Gerstner and Day’s (1997) LMX meta-analysis, the quality of the relationship between supervisor-subordinate was assessed by both supervisors and subordinates using the LMX7 measure (Graen & Uhl-Bien, 1995). The central item in this measure is “How effective is your working relationship with your leader?” Items were assessed on several different 5-point scales (see Appendices H and I for scale anchors). Coefficient alpha reliabilities for supervisor- and subordinate-rated LMX were .85 and .94, respectively (.92 overall). Interest in leadership development. Both supervisors and subordinates provided selfassessments of their interest in leadership development. Items assessed interest in participating in a leadership development program, searching for leadership development, and a desire to develop as a leader. Items were rated on a 7-point scale from strongly disagree (1) to strongly agree (7). All items loaded highly on a single factor (principal axis factoring with varimax rotation), which accounted for 89% of the variance in items. Coefficient alpha reliabilities for supervisor and subordinate data were .89 and .94, respectively (.94 overall). Behavioral repertoire of supervisors. Subordinate assessments of their supervisor’s behavioral repertoire was measured using Denison, Hooijberg, and Quinn’s (1995) measure of competing leadership behaviors. The four clusters of “competing” behaviors fall along an internal/external axis and a flexibility/control axis, and are labeled people, adaptive, stability, and task functions. Coefficient alphas were: .95, .96, .93, and .97, respectively. Subordinates were asked to assess their supervisor’s behaviors for each of the 16 items on a 7-point frequency scale, from almost never (1) to almost always (7). Self-confidence in leadership (leadership self-efficacy). Subordinate self-confidence in leadership was assessed using 7 items, based on measures developed by Feasel (1995) and Chan and Drasgow (2001). Items were assessed on a 1-7 scale, with confidence anchors
69 ranging from “0% confidence” (1) to “100% confidence” (7). A sample item is “I feel confident that I can lead others effectively”. Principal axis factor analysis (with varimax rotation) found that all seven items loaded highly on one factor (interpreted as a selfconfidence in leadership factor). This factor accounted for 82% of the variance among items. Coefficient alpha reliability for this scale was .94. The full-version of the scale can be found in Appendix I. Supervisor-rated leadership potential of subordinates. Leadership potential of each subordinate was assessed by the direct supervisor using a four-item scale developed for this study. The scale centered around the item “How much confidence do you have in his/her (your leader’s) potential to learn to be an effective leader?” Each item was assessed on an 11-point scale, ranging from 0% confidence (0) to 100% confidence (10) along an increasing continuum of 10% intervals. A full list of the four items can be found in Appendix H. Exploratory factor analysis (principal axis factoring with varimax rotation) suggested that one leadership potential factor explained a majority of the variance (88%) in the items. Coefficient alpha reliability was .96.
Results Hypotheses 7 through 14 were assessed in Study 2. Hypothesis 7 predicted mean differences between supervisors (physician managers and nurse managers) and subordinates (physicians and nurses) on individualism/collectivism dimensions, although these data were not collected due to survey length constraints. Instead, exploratory analyses examined mean differences between supervisors and subordinates on measures of leadership self-identity, orientation toward leadership, and interest in leadership development. Mean differences between supervisors and subordinates were compared using the t-statistic. Even though group sizes were quite different (~44 vs. 183), the t-statistic is relatively robust to violations of the equal sample size assumption, especially when both group sizes are above 30 (Hays, 1994). As Table 11 shows, supervisors and subordinates differed on several dimensions, including the descriptiveness dimension of self-identity; 4.63 vs. 4.24, t (225) = -1.80, p < .10, and the composite Overall Leadership Self-Identity; 4.49 vs. 4.16, t (225) = -1.68, p < .10. For both of these measures, supervisors had a stronger reported leadership self-identity. Supervisors and subordinates did not differ on the certainty and importance dimensions of self-identity. Orientation toward leadership of supervisors and subordinates did not differ for the shared and development dimensions, however, supervisors were less likely to endorse a dominance approach to leadership; 2.15 vs. 2.56, t (85) = 2.94, p < .01 (equal variances not assumed). Finally, supervisors reported higher interest in developing leadership skills and potential than did subordinates; 5.71 vs. 4.99; t (224) = 2.74, p < .01. In addition to examining mean differences between supervisors and subordinates, overall correlations between all measured variables are presented for the full sample (Table
70 12), as are correlations between measured variables, separated by supervisor and subordinate subsamples (Tables 13 and 14 respectively).
Table 11 Supervisor and Subordinate Means on Leadership Self-Identity, Orientation Toward Leadership and Leadership Development Items
†
Variable Leadership Self-Identity (Descriptive)
s.d. 1.34 1.17
df
T
Subordinate Supervisor
N mean 183 4.24 44 4.63
225
-1.80†
Leadership Self-Identity (Certainty)
Subordinate Supervisor
183 44
5.28 5.56
1.39 1.10
79a
-1.42
Leadership Self-Identity (Importance)
Subordinate Supervisor
183 44
4.09 4.35
1.52 1.53
225
-1.00
Leadership Self-Identity Core Factorb
Subordinate Supervisor
183 44
4.16 4.49
1.28 1.20
225
-1.54
Leadership Self-Identity (Overall Mean)
Subordinate Supervisor
183 44
4.54 4.85
1.13 .98
225
-1.68†
Orientation Toward Leadership (Shared)
Subordinate Supervisor
187 44
4.55 4.70
.98 .91
229
-.93
Orientation Toward Leadership (Dominance)
Subordinate Supervisor
187 44
2.56 2.15
1.06 .78
85 a
2.94**
Orientation Toward Leadership (Development/Influence)
Subordinate Supervisor
187 44
5.95 6.01
.81 .62
229
-.45
Participate in Leadership Development
Subordinate Supervisor
183 44
4.99 5.71
1.60 1.39
225
-2.74**
p < .10. ** p < .01. assumption of equality of variance not met (p < .05); therefore t-test calculated on adjusted degrees of freedom using Sattherwaite approximation. b Core Leadership Self-Identity is composed of all descriptiveness and importance items loading on highly on the first factor in a principal axis factor analysis.
a
Table 12 Correlations Among Common-Measured Variables for Medical Center Sample (Supervisors and Subordinates)
1 2 3 4 5 6 7 8 9 10 11 12
Scale
M
SD
1
2
3
4
5
6
Age Tenure LSI descriptiveness LSI certainty LSI importance LSI Factor Overall (grand mean) LSI OTL shared OTL Dominance OTL Development Leader-member exchange Interest in leadership dev.
3.10
1.05
--
9.29
8.41
.48
--
4.31
1.32
.02
.00
.87
5.34
1.34
.01
.11
.38
.95
4.14
1.52
-.06
-.09
.59
.35
4.23
1.27
-.02
-.05
.87
4.60
1.11
-.01
.00
.82
4.58
.97
.05
.15
.41
.93 .91
.91
.72
.84
.93
.90
-.01 -.01
-.05
.84
2.48
1.03
-.05
-.11
5.96
.77
.06
-.01
3.60
.99
.17
.15
.05
.12
5.14
1.58
-.04
-.21
.58
.18
-.02 -.09 -.05
7
8
9
.09
.08
.03
.05
-.06
.70
.08 -.11
.07
.08
.02
.22
-.21
-.12 -.05
.02
-.10
.55
.06
.53
.62
10
12
.92 -.02
.94
.77
-.03 -.10 -.03
11
.24
Correlations at or above .11, p < .10. Correlations at or above .14, p < .05. Correlations at or above .17, p < .01. N = 213 – 227, depending on specific correlations. Age was assessed on a 1-5 scale: 1= 20-30 years, 2 = 31-40 years, 3 = 41- 50 years, 4 = 51- years, 5 = over 60 years. LSI = Leadership self-identity.
71
Table 13 Correlations Among All Supervisor Assessed Variables (Medical Center Sample) Scale
1 2 3 4 5 6 7 8 9 10 11 12 13
Age Tenure (yrs.) LSI descriptiveness LSI certainty LSI importance LSI Factor Overall (grand mean) LSI OTL shared OTL Dominance OTL Development Leader-member exchange Interest in leadership dev. Subordinate ldrshp. potential
M
SD
1
2
3
4
5
6
3.51
0.75
--
12.75
9.93
.40
--
4.63
1.17
.03
-.06
.80
5.56
1.10
-.14
.30
.29
.90
4.35
1.53
-.11
-.19
.56
.30 .33
.91 .91
4.49
1.20
-.05
-.15
.85
.89
4.85
0.98
-.09
.00
.80
.64
.85
.94
4.70
0.91
.11
.18
-.22
-.06
-.12
2.15
0.78
.13
.00
-.13
.19
.21
6.01
0.62
.38
.19
.13
-.18
4.02
0.52
.20
.20
-.14
5.71
1.39
-.06
-.34
6.24
1.57
.16
.23
7
9
10
12
13
.83
.13
.22
.55
-.10
.00 -.07
.25
-.07
.69
.04
-.26
-.24 -.18
-.12
.07
-.31
.42
-.05
.37
.34
.20
-.12
-.24
-.09
-.22
-.26 -.25
-.08
.15
.45
11
.88
-.18 -.17 .07
8
.85
.31 -.45 .01
.57
.89 -.39
.96
Correlations at or above .26, p < .10. Correlations at or above .31, p < .05. Correlations at or above .38, p < .01. N = 41 - 44, depending on specific correlations. Age was assessed on a 1-5 scale: 1= 20-30 years, 2 = 31-40 years, 3 = 41-50 years, 4 = 51- years, 5 = over 60 years. LSI = Leadership self-identity. OTL = Orientation toward leadership.
72
Table 14 Correlations Among All Subordinate Assessed Variables (Medical Center Sample) Scale
M
SD
1
1 Age
3.00 1.09
--
2 Tenure (yrs.)
8.49 7.83
.48
3 LSI descriptiveness
4.23 1.34
.00 -.01
.88
4 LSI certainty
5.28 1.39
.01
.05
.39
.96
5 LSI importance
4.09 1.52 -.06 -.08
.59
.36
.93
6 LSI Factor
4.16 1.28 -.03 -.05
.88
.42
.91
.92
7 Overall (grand mean) LSI
4.54 1.13 -.02 -.02
.82
.73
.83
.93
8 LWV shared
4.55 0.98
.01 -.11
.01
.01 -.03
9 LWV Dominance
2.56 1.06 -.04 -.10 -.01
.09
.08
.04
.07 -.10
.10
.09
.03
.03
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
--
.14
.92 .84 .73
10 LWV Development
5.95 0.81
.01 -.07
.06 -.10
11 Leader-member exchange
3.49 1.04
.14
.11
.04
.11 -.12 -.05
.01 -.11
.00 -.09
12 Interest in leadership dev.
4.99 1.60 -.09 -.24
.60
.20
.57
.02
13 Behavioral repertoire (people)
4.84 1.75
.02
14 Behavioral repertoire (adaptive)
4.39 1.75
15 Behavioral repertoire (stability)
.04 -.01
.56
.65
.21 -.23 .02
.79 .94
.23 -.03
.94
.12 -.15 -.10 -.02 -.11 -.01 -.07
.86 -.03
.95
.04 -.07 -.01
.12 -.14 -.09 -.02 -.08
.08 -.03
.79 -.03
.84
.96
4.39 1.62
.05
.05 -.16 -.14 -.08 -.11
.15 -.14
.80 -.06
.82
.83
.93
16 Behavioral repertoire (task)
4.92 1.68
.06 -.03
.20 -.07 -.03
.06 -.06
.07 -.10
.80 -.03
.83
.81
.88
.97
17 Behavioral repertoire Overall
4.63 1.60
.04 -.01 -.02
.13 -.14 -.10 -.02 -.10
.07 -.09
.87 -.04
.93
.93
.94
.94
.97
18 Confidence in ldrship abilities
5.06 1.27
.07 -.01
.35
.01 -.01
.08
.01
.01 -.08 .02 .61
.39
.55
.56 -.09 -.06 -.04
.09
.49 -.02
73
Correlations at or above .13, p < .10. Correlations at or above .15, p < .05. Correlations at or above .19, p < .01. N = 176 - 183, depending on specific correlations. Age was assessed on a 1-5 scale: 1= 20-30 years, 2 = 31-40 years, 3 = 41-50 years, 4 = 51- years, 5 = over 60 years. LSI = Leadership self-identity.
.94
74 Testing Hypotheses 8, 9, and 10 Using Hierarchical (Multi-level) Models Hypotheses 8 and 9 predicted that leadership self-identity of supervisors would be negatively related to mean supervisor assessments of leader-member relationship quality (LMX; H8) and subordinate leadership potential (H9). The main effects of supervisors’ orientation toward leadership may also relate to supervisor’s giving lower LMX and potential ratings. The latter are considered exploratory analyses. Finally, Hypothesis 10 predicted an interaction between leadership self-identity and orientation toward leadership in predicting LMX and leadership potential ratings. Two possibilities exist in analyzing these hypotheses. First, these questions might be answered using correlations/regression analyses and, second, they can be answered with Hierarchical Linear Modeling (HLM) if data are non-independent and the outcome of interest is at the lower-level of analysis. In the present study, leaders rated each subordinate on two variables: (a) subordinate leadership potential, and (b) leader’s relationship with subordinate. Because each supervisor provided ratings of subordinates, these data are non-independent, or nested (see Figure 8). The fact that data are nested and our outcome exists at the individual-level of analysis allows for the possibility of hierarchical modeling. In HLM terms, individuals (and ratings of individuals) are at the lower level of analysis, and nested within leaders (including leader self-identity and orientation toward leadership; see Fig. 8). HLM overcomes the disadvantages of non-hierarchical approaches (e.g., regression) because one is able to explicitly model both within and between group variance simultaneously, as well as investigate the influence of higher level units on lower level outcomes while maintaining the appropriate level of analysis (i.e., HLM recognizes the fact that individuals within a group may be more similar than others outside the group; Hofmann, 1997). Stated slightly differently, regression assumes the same intercept and slope for all individuals, while HLM allows for a different slope and intercept for each participant at level 1. One of the practical results of using this approach is that, compared to non-nested approaches, HLM decreases the probability of incorrectly rejecting the null hypothesis at the higher level of analysis (Level 2 predictor), and failing to reject a true null hypothesis at the lower level of analysis (Level 1 predictor). Thus, in the case of Hypotheses 8 to 10, HLM is the appropriate data analytic technique.
75
Leader 1
Level 2
Sub11 Sub21 Sub31 Sub41 Sub51 Level 1
Leader 2
Sub12 Sub22 Sub32
Figure 8. Example of nested nature of supervisor-ratings of subordinate leadership potential.
As an example, in the case of supervisor ratings of subordinate leadership potential (Figure 8), the outcome that we are trying to predict is each individual’s ratings, or “scores”. There are probably individual factors that predict an individual’s score, which might include variables such as charisma, intelligence, or overall job performance of the individual subordinate. These individual level predictors, however, are not of interest in the present research. What is of interest is the extent to which there are characteristics of the leaders (i.e., leadership self-identity and orientation toward leadership) that affect the ratings they provide about their subordinates and the quality of relationships they have with them. In other words, individual ratings are not independent of the leader (Level-2) who is providing the ratings. Some leaders may give higher ratings than other leaders, and the extent to which a leader gives relatively higher or lower ratings is, according to the hypotheses in this research, at least partly based on leadership self-identity and orientation toward leadership. In the HLM equation, the leader effects are at Level-2, and we are trying to predict an individual’s rating (level 1) based on the leader’s characteristics (level 2). This “intercepts as outcome” model is the simplest HLM model in that individual ratings are estimated by the model intercept and error at level 1. The intercept estimates are then modeled with leader information at level 2. This model seeks to examine differences in ratings of supervisorrated potential and LMX according to the supervisor’s leadership self-identity. An example of the generic two-level model predicting supervisor ratings of employee leadership potential is presented below. It should also be noted that an identical two-level equation holds for the leadership orientation dimensions and the dependent variable of supervisor-rated LMX. In accordance with Hofmann, Griffin, & Gavin’s (2000) suggestion of not using group mean centered variables in intercepts-as-outcomes models (where predictors for each group are centered around each group’s mean), raw metric level-2 predictors (leadership self-identity and orientation toward leadership) were used. Alternatively, predictors could have been centered around the overall mean across groups (grand mean centering), though grand mean centering has been shown to be equivalent to raw metric scaling (Kreft, DeLeeuw, & Aiken, 1995).
76 This model takes the form of a 2-level equation such that: Level 1 Leadership Potential Rating = B0 + r Where: B0 = Level-1 intercept (random coefficient) r = Level-1 within-group error
Level 2 B0 = γ00 + γ01*(LSI Dimension) + U0. Where: B0 = Level-1 intercept γ00 = Level-2 intercept γ01*(LSI Dimension) = Level-2 slope (grand mean centered LSI dimension) U0 = between-group error in intercepts
Before assessing the substantive models with the Level-2 predictors, it is useful to examine the between-group variance as a proportion of total variance in the null model (i.e., should the intercept be considered a random effect). The ratio of between-group variance to total variance is also known as an intraclass correlation (ICC; Hofmann, Griffin, & Gavin, 2000) and is calculated from the unrestricted model as: between-group variance (between-group variance + within-group variance) For the dependent variable LMX with no predictors (unrestricted model), obtained ICC values of .49 (=.184/[.184+.191]) indicate that 49% of the variance in LMX ratings resides between groups. This non-trivial amount of group dependence in LMX ratings suggests that there are considerable group-level effects that should be examined. For the dependent variable of leadership potential ratings, obtained ICC values of .14 (=.697 / [.697+4.256]) indicate that 14% of the variance in leadership potential ratings resides between groups. This value, although lower than that found for LMX ratings, is still indicative of significant group effects (Bliese, 2000). As a further rigorous assessment of the significance of group-level effects, it may be desirable to compute log likelihood tests to contrast fixed versus random intercept models. However, Bliese and Ployhart (2002, p. 380) suggest that this test is necessary only when ICC values are less than .10. Values above .10 indicate sufficient evidence for group-level effects; therefore log likelihood contrasts are not examined. Because little is known about leadership self-identity and its relationship to possible outcomes, four different models were assessed to test Hypothesis 8; three models with one of
77 the three self-identity dimensions, and one model with all three self-identity dimensions together. In all HLM models (see Tables 15 to 18) Leadership Self-Identity dimensions do not predict LMX (as evidenced by non-significant t-tests of predictors). This effect was consistent, regardless of whether the Leadership Self-Identity dimensions were considered separately or together. Further, although the test is imperfect (Bliese & Ployhart, 2002), the significant chi-square test of residual intercept variance (Level-1) is suggestive that systematic Level-2 variance remains that could be modeled by additional level 2 predictors (Hofmann, Griffin, & Gavin, 2000). In sum, the consistent patterns indicate that there is significant variability between leaders in ratings of LMX (as indicated by substantial ICC values), but Leadership SelfIdentity dimensions do not explain that variability. One specific result worthy of mention, however, is that the coefficient for Leadership Self-Identity Importance in predicting LMX approached a relaxed significance level in the combined model (p = .131). This result, although interesting at this exploratory phase, should be treated with considerable caution. Hypothesis 8 predicting that leadership self-identity is negatively related to supervisor-rated LMX is not supported.
Table 15 Hierarchical Linear Model (HLM) Results Predicting Supervisor Ratings of Subordinate-Level LMX with Leadership Self-Identity Descriptiveness Level 1 Variance Components Intercept 1 Level-1
Standard Deviation
Variance Component
.43 .44
.19 .19
Coefficient (unstandardized)
Standard Error
Chi Square 286.02**
Level 2 Fixed Effect For Intercept 1, B0 Intercept 2 LSI: Descriptiveness ** p < .01, df = 42.
4.01 -.04
.07 .06
t-ratio
54.99** -.06
78 Table 16 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate-Level LMX with Leadership Identity Certainty Level 1 Variance Components
Standard Deviation
Variance Component
Chi Square 284.77**
Intercept 1
.43
.19
Level-1
.45
.19
Level 2 Fixed Effect
Coefficient (unstandardized)
Standard Error
t-ratio
.07 .06
54.96** .34
For Intercept 1, B0 Intercept 2 LSI: Certainty ** p < .01, df = 42.
4.01 .02
Table 17 Hierarchical linear model (HLM) Results: Predicting Supervisor Ratings of Subordinate-Level LMX with Leadership Identity Importance Level 1 Variance Components Intercept 1 Level-1
Standard Deviation .42
Variance Component .18
.44
.19
Coefficient (unstandardized)
Standard Error
4.01 -.08
.07 .06
Chi Square 271.47**
Level 2 Fixed Effect
t-ratio
For Intercept 1, B0 Intercept 2 LSI: Importance ** p < .01, df = 42.
56.12** 1.35
79
Table 18 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate-Level LMX with All Three Leadership Identity Dimensions Level 1 Variance Components Intercept 1 Level-1
Standard Deviation .43 .44
Variance Component .18
Chi Square 261.94**
.19
Level 2 Fixed Effect
Coefficient (unstandardized)
Standard Error
t-ratio
For Intercept 1, B0 Intercept 2 LSI: Descriptiveness LSI: Certainty LSI: Importance ** p < .01, df = 40.
4.01 .02 .05 -.09
.07 .06 .05 .06
56.91** .29 1.00 -1.65
In order to test Hypothesis 9, which stated that supervisor leadership self-identity is related to lower ratings of subordinate leadership potential, HLM was used in a manner similar to tests of Hypothesis 8. Results of the analyses predicting leadership potential with individual supervisor Leadership Self-Identity dimensions are presented in Tables 19 through 21. Results of the hierarchical model predicting leadership potential with all three Leadership Self-Identity dimensions are presented in Table 22. In all HLM models, supervisor Leadership Self-Identity does not account for significant variability in ratings (as evidenced by non-significant t-tests). Similar to tests of Hypothesis 8, the significant chisquare test of residual intercept variance (Level-1) in the four models indicates that there remains systematic Level-2 variance that could be modeled by additional level 2 predictors. In sum, Hypothesis 9, that leadership self-identity is negatively related to supervisor-rated LMX, is not supported.
80 Table 19 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate Leadership Potential with Leadership Identity Descriptiveness Level 1 Variance Components Intercept 1 Level-1
Standard Deviation
Variance Component
.85
.73
2.12
4.51
Chi Square 79.41**
Level 2 Fixed Effect For Intercept 1, B0 Intercept 2 LSI: Descriptiveness
Coefficient (unstandardized) 6.14 -.14
Standard Error .20 .17
t-ratio
30.02** -.84
** p < .01, df = 42. Table 20 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate Leadership Potential with Leadership Identity Certainty Level 1 Variance Components Intercept 1 Level-1
Standard Deviation
Variance Component
.82
.67
2.13
4.54
Chi Square 76.56**
Level 2 Fixed Effect For Intercept 1, B0 Intercept 2 LSI: Certainty ** p < .01, df = 42.
Coefficient (unstandardized) 6.12 .14
Standard Error .19 .16
t-ratio
31.94** .85
81
Table 21 Hierarchical Linear Model (HLM) Results: Predicting Supervisor-Ratings of Subordinate Leadership Potential with Leadership Identity Importance Level 1 Variance Components Intercept 1 Level-1
Standard Deviation .86
Variance Component .73
2.12
4.51
Coefficient (unstandardized)
Standard Error
6.13 -.10
.20 .14
Chi Square 79.82**
Level 2 Fixed Effect For Intercept 1, B0 Intercept 2 LSI: Importance ** p < .01, df = 42.
t-ratio
30.23** -.71
82 Table 22 Hierarchical Linear Model (HLM) Results: Predicting Supervisor Ratings of Subordinate Leadership Potential with All Three Leadership Identity Dimensions Level 1 Variance Components Intercept 1 Level-1
Standard Deviation .85
Variance Component .72
Chi Square 72.40**
2.13
4.53
Coefficient (unstandardized)
Standard Error
t-ratio
6.15 -.13 .19 -.09
.20 .21 .18 .16
30.25** -.62 1.06 -.52
Level 2 Fixed Effect For Intercept 1, B0 Intercept 2 LSI: Descriptiveness LSI: Certainty LSI: Importance ** p < .01, df = 40.
Although not explicitly hypothesized, main effects of orientation toward leadership dimensions were also assessed in a series of exploratory analyses to determine whether supervisors holding certain theories of leadership rated subordinate LMX and leadership potential differently. From an assessment of the null models (described previously), it is known that between-group variance accounts for a significant proportion of total variance in LMX (49%) and leadership potential (14%) ratings, suggesting that there are “leader-level” effects. Hierarchical models with leadership orientation showed a pattern: Orientation Toward Leadership dimensions did not predict LMX or subordinate leadership potential. One exception to this general trend, however, is the effect of development orientation on ratings of LMX (see Table 23). Interestingly, supervisors with a more developmental view of leadership rated subordinate LMX as lower, t (42) = -2.16, p < .05. Hypothesis 10 posited interactions between a supervisor’s orientation toward leadership and self-identity in predicting both LMX ratings (H10a) and supervisor ratings of subordinate leadership potential (H10b). Leadership self-identity is believed to be negatively related to supervisor ratings, particularly when the content of a supervisor’s leadership orientation is high on dominance and low on Development and shared dimensions. A series of 18 HLM models were analyzed (3 leadership orientations x 3 self-identity dimensions x 2 dependent variables).
83 Table 23 Hierarchical Linear Model (HLM) Results: Predicting Leader-Member Exchange Quality with Development Orientation Toward Leadership Level 1 Variance Components
Standard Deviation
Variance Component
Chi Square
Intercept 1
.41
.17
Level-1
.44
.19
Coefficient (unstandardized)
Standard Error
t-ratio
4.01 -.25
.07 .12
57.23** -2.16*
264.72**
Level 2 Fixed Effect For Intercept 1, B0 Intercept 2 OTL: Development ** p < .01
* p < .05,
df = 42.
None of the interactions predicting LMX with leadership self-identity and leadership orientations were significant; thus Hypothesis 10a is not supported. Several interactions predicting leadership potential (Hypothesis 10b), however, were significant. Tables 24 and 25 present HLM interactions between leadership identity and dominance orientation in predicting supervisor ratings of leadership potential. In order to better understand these interactions, Figures 9 and 10 graphically depict the relationships. As can be seen in both figures, having an orientation toward leadership that is high in dominance actually leads to higher ratings of subordinate leadership potential. The line representing low Dominance (1 standard deviation below mean) shows that leadership potential is rated lower, regardless of Leadership Self-Identity. However, supervisors with both a low dominance view and a low self-identity were predicted to rate subordinates the lowest, of all four points depicted. Taken together, although leadership orientation and leadership self-identity interacted in predicting supervisor ratings of leadership potential in several of the interactions tested, the shape of the interaction was different from that hypothesized; thus Hypothesis 10b was not supported.
84 Table 24 Supervisor Ratings of Subordinate Leadership Potential with LSI Descriptiveness and Dominance Orientation Level 1 Variance Components Intercept 1 Level-1
Standard Deviation .75
Variance Component .57
2.13
4.52
Coefficient (unstandardized)
Standard Error
Chi Square 67.34**
Level 2 Fixed Effect
t-ratio
For Intercept 1, B0 .19 .61 1.33 .26 df = 40.
32.40** 1.66† 2.12* -1.91†
10 9 Sub Leadership Potential
†
Intercept 2 6.11 LSI: Descriptiveness 1.02 OTL: Dominance 2.82 Desc. X Dominance -.50 p < .10 ** p < .05 ** p < .01.
8 7 6 Low Dominance High Dominance
5 4 3 2 1 0 Low
High
LSI Descriptiveness
Figure 9. Subordinate leadership potential by supervisor leadership identity descriptiveness.
85 Table 25 Supervisor Ratings of Subordinate Leadership Potential with LSI Importance and Dominance Orientation Level 1 Variance Components
Standard Deviation .71
Intercept 1 Level-1
2.13
Variance Component .51
Chi Square 64.01**
4.53
Level 2 Fixed Effect
Coefficient (unstandardized)
Standard Error
t-ratio
For Intercept 1, B0 6.15 .70 2.23 -.41 ** p < .01.
.18 .46 .89 .20 df = 40.
34.72** 1.53 2.50* -2.00*
9 8 Sub. Leadership Potential
†
Intercept 2 LSI: Importance OTL: Dominance Importance x Dom. p < .10 ** p < .05
7 6 5
Low Dominance
4
High Dominance
3 2 1 0 Low
High LSI Importance
Figure 10. Subordinate leadership potential by supervisor self-identity importance.
86 Hypothesis 11-14 Hypothesis 11 posited a positive relationship between a supervisor’s leadership selfidentity and subordinate assessment of the supervisor’s behavioral repertoire. To calculate subordinate ratings in the case of multiple subordinates providing ratings of a single supervisor, mean ratings were calculated. As can be seen from Table 26 (top half of table), none of the Leadership Self-Identity dimensions were positively related to behavioral repertoire. One Leadership Self-Identity dimension was marginally (p = .10) related to one of the dimensions of behavioral repertoire in the negative direction. However, because of the number of correlations that were calculated, this result should be interpreted with extreme caution as it may be due to chance. Although the relationships were almost wholly insignificant, it should be noted that the trend in the direction of these relationships is negative, which is contrary to the hypotheses. Twenty-two of the twenty-five correlations were negative in sign. Thus, Hypothesis 11 predicting that leadership self-identity of supervisors is positively related to subordinate ratings of behavioral repertoire is not supported. Significant relationships between leadership orientation and subordinate ratings of supervisor behavioral repertoire were also hypothesized (H12), such that supervisors with higher shared and developmental views of leadership were expected to have a greater behavioral repertoire. None of these correlations were significant (lower half of Table 26). Thus, Hypothesis 12 is not supported. Leadership views of the supervisor are not related to subordinate assessments of behavioral repertoire. Although not explicitly hypothesized, exploratory correlations between subordinaterated LMX and supervisor’s Leadership Self-Identity and leadership orientation were examined. The importance dimension of self-identity was negatively related to relationship quality (LMX), such that holding leadership identity as important to the self was negatively related to subordinate assessments of relationship quality, r(38) = -.36, p < .05. None of the other component dimensions were significantly related, nor were supervisor Orientation Toward Leadership dimensions. Thus, these results are not reported.
87 Table 26 Correlations Between Leadership Self-Identity of Supervisor and Subordinate Ratings of Supervisor Behavioral Repertoire People Behavior
Adaptiveness Stability Behavior Behavior
Task Behavior
Overall Behavioral Repertoire
Descriptiveness LSI
-.06
.03
-.03
.06
.00
Certainty LSI
-.27
-.17
-.12
-.12
-.20
Importance LSI
-.28†
-.24
-.10
-.09
-.20
LSI Factor (Desc. + Importance)
-.21
-.14
-.08
-.03
-.12
Overall LSI
-.27
-.18
-.11
-.07
-.18
Dominance Leadership View
-.18
.05
.04
-.03
-.04
-.04
-.03
-.09
-.09
-.07
.00
-.03
.02
.00
.00
Development Leadership View Shared Leadership View † significant at p < .10
Minimum N = 35.
Hypothesis 13a predicted that leadership self-identity is related to personal interest in leadership development. Interest in leadership development was assessed among both supervisors and subordinates, and significant positive correlations were observed with all Leadership Self-Identity dimensions and interest in development (see Table 12). Of the Leadership Self-Identity dimensions, correlations ranged from .18 to .58; r(230), all p < .01. When the dimensions were assessed separately for both the supervisor and subordinate samples (Tables 13 and 14), all correlations were significant and positive (all p < .01), except for the relationship between certainty and interest in development in the supervisor sample, r(42) = -.05, p > .10. Taken together, hypothesis 13a is supported; leadership self-identity is related to interest in leadership development. Hypothesis 13b suggested that views about leadership are related to interest in leadership development. The development view of leadership, which gauges the extent to which one believes that leadership can be developed, is likely to be most related to interest in leadership development. Results supported this hypothesis for the Development dimension
88 of Orientation Toward Leadership, r(42) = .24, p < .01. Shared views of leadership and dominance views were not related to interest in leadership development (p > .05). Hypothesis 13b is supported, beliefs that leadership can be developed is related to interest in leadership development. Hypothesis 14 posited an interaction between leadership self-identity and views about leadership in predicting willingness to engage in leadership development. More specifically, leadership self-identity should be more positively related to interest in leadership development when an individual’s leadership orientation is low on personal dominance and high on Development/Influence and Shared orientations. A series of nine moderated regressions were calculated, with each of the three specific Leadership Self-Identity dimensions crossed with each of the three dimensions of leadership views. A Leadership Self-Identity dimension and a leadership view dimension were entered in the first step of the regression (mean centered), followed by the interaction term in the second step. Out of nine possible interactions, two were significant. Results are presented in Table 27 and a graphic representation of these interactions follows in Figures 11 and 12. Consistent with the general hypothesis, the interaction between self-identity importance and shared orientation (Figure 11) was such that self-identity was positively related to interest in development for those high in shared orientation toward leadership, but negatively related to interest in leadership development for those low in shared orientation. Having a high importance and low shared approach, or a low importance and high shared view of leadership, resulted in lower reported interest in leadership development. The interaction between certainty of self-identity and dominance orientation was such that the positive relationship between self-identity and interest in development was stronger when an individual had a low dominance orientation. Consistent with the general hypothesis for high dominance orientation individuals, self-identity certainty had little effect on interest in leadership development. None of the interactions involving development orientation toward leadership and self-identity were significant. Overall, there is some support for Hypothesis 14.
89 Table 27 Summary of Significant Moderated Regressions: Predicting Interest in Leadership Development with Leadership Self-Identity and Orientation Toward Leadership Predictor Step 1 LSI Importance OTL Shared R2 Adjusted R2 Fb Step 2 LSI Importance OTL Shared Importance x Shared
.532** .061
Predictor Step 1 LSI Certainty OTL Dominance R2 Adjusted R2 Fb
.29** .28** 44.84** .07 -.22 .55*
Step 2 LSI Certainty OTL Dominance Certainty x Dominance
.01 R2 Change Total R2 .30** Adjusted R2 .29** Fb 31.48** † p < .10 ** p < .05 ** p < .01. a standardized regression coefficients are shown. b df = 3, 223.
R2 Change Total R2 Adjusted R2 Fb
-.649 2.762** .03† .03† 3.90* 1.69† 2.83** -1.89† .02† .05* .04* 3.82**
90
Interest in Leadership Development
6.5
6
5.5 Low Shared High Shared 5
4.5
4 Low
High LSI Importance
Figure 11. Interest in leadership development by leadership identity importance (supervisors and subordinates).
91
Interest in Leadership Development
6.5
6
5.5 Low Dominance High Dominance
5
4.5
4 Low
High LSI Certainty
Figure 12. Interest in leadership development by leadership identity certainty (supervisors and subordinates).
92 Discussion Overview The primary purpose of Study 2 was to examine leadership self-identity and orientation toward leadership of both supervisors and subordinates in a field setting. Differences between supervisors and subordinates were examined, as were outcomes that may be associated with both self-identity for leadership and leadership orientation. A general overview is presented in this section, followed by a more detailed discussion in the following section. Hypothesis 7 was not directly tested. After discussion with the Chief of Human Resources of the hospital, a decision was made that the proposed survey needed to be shortened to increase response rate and reduce the time burden placed on participants. Given that individualism and collectivism were assessed in Study 1 and that this measure was fairly long (27 items total), the measure of individualism and collectivism was dropped. Selfidentity and leadership orientation were compared in supervisor and subordinate samples in an extension of Hypothesis 7. In general, supervisors had marginally higher self-identity descriptiveness than subordinates, but supervisors and subordinates reported certainty and importance of leadership self-views similarly. Supervisors were less endorsing of a dominance approach to leadership and were more likely to report interest in leadership development. Orientation toward leadership and self-identity may be reliably measured and related to other personality/individual difference variables, but if there is no evidence of these constructs being related to relevant outcomes, then the case for their importance as constructs is less established. Overall, there was some support for the relevance of these dimensions, though a number of hypotheses were not supported. Leadership self-identity and leadership orientation generally did not have an effect on supervisor ratings of LMX with subordinates or subordinate leadership potential (Hypotheses 8, 9, and 10). Significant between-leader variance suggests that there are other leader or group-level variables that may account for this variance. Self-identity and leadership orientation of supervisors similarly did not predict subordinate ratings of the behavioral repertoire of the supervisor (Hypotheses 11 and 12). Self-identity (Hypothesis 13a) and leadership orientation (Hypothesis 13b) were predictive of interest in leadership development, and in two cases interacted in predicting interest in development (Hypotheses 14). Differences Between Subordinates and Supervisors It is not simply supervisors or those in formal leadership positions who have a leadership self-identity and hold beliefs about leadership; all individuals hold opinions about leadership and about themselves as leaders (or not leaders). With this in mind, exploratory comparisons of leadership identity and leadership orientation for both supervisors and subordinates were undertaken. At a relaxed significance level (p < .10), there was a
93 difference between supervisors and subordinates in Leadership Self-Identity Descriptiveness, with supervisors reporting that the leadership identity items were more descriptive of themselves. For the certainty and importance dimensions, however, there were no differences. In sum, the differences in self-identity were either nonexistent or small, suggesting that while supervisors in this sample may have slightly higher perceptions of themselves as leaders, these differences are not large. One explanation for this finding may be that individuals play multiple roles (Merton, 1957), both within and outside the work domain, and these roles may involve leadership. For example, a subordinate nurse may help to run a committee at work, or play a leadership role in a religious organization outside of work. Leadership is all around us (Yukl, 2002), it can take many forms (e.g., Hiller, Day, and Vance, in press), and many individuals may be involved in leadership at some level, even though their primary work role may not be one of formal authority. Regardless of the reasons, the general lack of supervisor/subordinate differences is suggestive that possession of a formal leadership role at work is not required to have an identity for leadership. There was however a difference between supervisors and subordinates in the extent to which they held a dominance leadership orientation. Supervisors, who by nature of their position possess formal authority, were less endorsing than subordinates of a dominance approach to leadership that, in part, emphasizes formal authority. Perhaps their lower level of dominance orientation came through in their behaviors and is a reason for being promoted to a supervisory position. Generally, individuals high in the personality trait of dominance emerge as leaders (e.g., Smith & Foti, 1998), and so one might presume that the formal supervisors would have a higher dominance orientation. However, having a dominant personality is not the same as the belief that leadership can involve control and authority (i.e., dominance leadership orientation), although they are likely related. There were no differences between supervisors and subordinates in the shared and development dimensions of Orientation Toward Leadership. Effects of Supervisor Leadership Self-Identity and Orientation Toward Leadership on LMX and Leadership Potential Ratings The single previous assessment of leadership perceptions of the self undertaken by Engle and Lord (1997) found that supervisory leadership self-schemas (the belief that one had traits that were generally endorsed as also being possessed by leaders) were related to poorer quality LMX relationships. One suggestion offered for this finding that is potentially congruous with the general literature on self-schemas was that schematic supervisors (those with a highly salient leadership self-identity) were more stringent judges of others in that domain. In the present study this finding was explored and extended using a more direct assessment of leadership self-identity, and both LMX and leadership potential ratings. If indeed strong leadership self-identity is related to stringency in ratings, we might expect an even greater relationship with ratings of leadership potential than relationship quality, since it is more directly a rating of subordinates in the task (leadership) domain than is LMX. In addition to these differences from the previous study, the present study used an advanced modeling technique for examining the non-independence of supervisor ratings (i.e., Hierarchical Linear Modeling or HLM).
94 Intraclass correlation coefficients (ICC1) in the present research suggest that different leaders provide ratings that are systematically different; however, seeing the self as a leader (as assessed by the Leadership Self-Identity measure) is not responsible for this systematic difference. The evidence from these hierarchical linear models ran counter to the hypotheses, and is generally inconsistent with the self-schema research from other domains. It may be, however, that there are opposing forces in operation causing the null findings. At the same time that there is a stringency effect such that high self-identity pushes people to lower assessments of others, it might also be that a strong leadership self-identity allows individuals to be less threatened by the potential leadership of others (regardless of their orientation toward leadership). Positive self views can act as a buffer against defensiveness (Brockner, 1988), and positive self-views in the leadership domain may act as a buffer against defensiveness in rating others, or seeing others as a threat. Another possible explanation for the null findings is the organizational context of this study. The organization was atypical from most hospitals (and indeed most organizations) in that there was an in-house leadership development center, and leadership development at all levels was continually discussed and promoted. This is not to say that all individuals participating in the survey had taken part in leadership development, but the cultural emphasis on leadership development may have had an effect in the way supervisors thought about and engaged in leadership. To the extent that the culture emphasized leadership development at all levels, and supervisors were not threatened by the notion of subordinates being leaders, we might expect that the supervisor stringency effect may be nullified. Although these possibilities are intriguing, it must be acknowledged that they are also speculative. Engle and Lord’s analysis relied on correlations between leadership self-schema and LMX ratings and thus was not able to properly model variance at the supervisor level, making it more prone to a false rejection of the null hypothesis that supervisor schemas had an effect on LMX ratings. In other words, their finding of negative correlations may be an artifact of the inability to model both supervisor-level and individual-level variance in supervisory assessments. Because of the bias in falsely rejecting the null hypothesis that is inherent when multi-level models are not properly analyzed, it is possible that something other than self-schematicity accounted for variance between supervisors. In the present study, bivariate correlations suggest a negative trending relationship between self-identity and ratings of LMX and potential, similar to what was reported by Engle and Lord. Hierarchical models, however, revealed that leadership self-identity does not account for leader-level differences. Although there is no main effect of leadership self-identity on LMX and ratings of subordinate leadership potential, significant interactions were observed in several cases when self-identity is combined with leadership orientation. These interactions suggest that one’s cognitive view of leadership and his or her self-identity around leadership need to be examined together to understand how perceptions of, and actions toward, others are affected. These interactions occurred when predicting leadership potential with supervisor dominance orientation and the self-identity dimensions of importance and descriptiveness. This
95 interaction took the form such that the level of LSI descriptiveness and importance impacted ratings more for those who were low in Dominance (see Figures 9 and 10). Individuals high in dominance views were less affected (in terms of leadership potential ratings) by their leadership self-identity. Those supervisors who were relatively higher on dominance orientation gave the highest ratings of leadership potential. These supervisors were not endorsing of a dominance approach; one standard deviation above the mean on the Dominance Orientation measure was still below the midpoint of the scale, suggesting that “high dominance orientation” leaders would still generally not endorse dominance, but may be more prone to seeing it as a possible leadership strategy. Sometimes dominance is a necessary and effective approach (Drath, 2001) and these higher dominance leaders may be more likely to acknowledge this notion. Further, these leaders, who are less opposed to the idea of dominance, may be the most secure or comfortable in their leadership because they are more endorsing of a leadership idea that is not generally popular in most organizations. These secure individuals may in turn be less prone to the downward bias in ratings of leadership potential. This possibility, while intriguing, is highly speculative and must be tested by further research. The relationships between supervisor self-identity, orientation toward leadership and supervisor ratings were mostly not borne out in the data. Exploratory correlations of supervisors’ identity and orientation with mean subordinate LMX ratings resulted in a slightly different picture. The importance dimension of supervisor leadership self-identity was negatively related to relationship quality (LMX) as rated by subordinates, r(38) = -.36, p < .05. This finding, when considered alongside the hierarchical linear modeling findings, may indicate that a supervisor’s leadership self-identity does not lead to being a more stringent rater of subordinates from the supervisor’s point of view, but may in fact lead to poorer relationships with subordinates through some other mechanism. The fact that the common-source relationships (where supervisor provides all of the ratings) were not significant in the HLM models, but the different-sourced relationships (supervisor identity and orientation, subordinate LMX) were significant is noteworthy, and should be examined further in the future. The extent to which these findings are replicated in other samples and the degree to which there might be something about the importance dimension of selfidentity that is particularly relevant to LMX relationships could provide a more detailed understanding of leadership self-identity. Interest in Leadership Development In the full supervisor/subordinate combined sample, leadership self-identity of an individual was related to their interest in leadership development, consistent with self-schema research showing that self-schematicity is related to interest in learning and demonstrating performance in a given domain (Cross & Markus, 1994; Ng, 2005). In other words, this may be an indicator of the phenomenon of the “rich getting richer.” Those with a high selfidentity for leadership are more likely to engage in future leadership development (or at least they report more interest in developing as a leader). Those with a lower leadership selfidentity, who may perhaps need leadership development the most, may be less likely to look
96 for opportunities to develop as a leader. Based on results from the Study 1, it is known that seeing the self as a leader is highly related to past leadership experiences. It could be that a cyclical effect operates such that having leadership experiences leads to a leadership identity, which in turn leads to interest in leadership development. Susan Komives and colleagues (e.g., Komives, Casper, Longerbeam, Mainella, & Osteen, 2004) provide evidence for this participation – identity developmental linkage. They found that both leadership identity and beliefs about leadership change in college students as a result of experiences. This suggests that active engagement in leadership may be important in order to develop a leadership identity. Perhaps individuals with a low leadership identity can develop a stronger identity for leadership through concerted participation in leadership. Once begun, the reinforcing cycle of leadership identity – leadership development and enactment may continue. A belief that leadership can be taught and developed (development orientation) was similarly related to interest in leadership development for the combined supervisor/subordinate sample. Given this finding, it is possible that orientation toward leadership may be another avenue (in addition to self-identity changes) to enact interest and participation in leadership development. A development orientation, as a cognitive orientation, might potentially be malleable (Wiers, de Jong, Havermans & Jelicic, 2004), just as the working self-concept can be altered in leadership situations (Lord & Brown, 2004). Interestingly, there were several interactions in which leadership self-identity interacted with orientation toward leadership dimensions in predicting interest in leadership development for the supervisor/subordinate sample (Hypothesis 11). For example, selfidentity was positively related to interest in leadership development for those high in shared orientation, but negatively related to interest in leadership development for those low in shared orientation. The other significant interaction was between Self-Identity Certainty and Dominance Orientation, such that the positive relationship between self-identity and interest in development was stronger when an individual had a low dominance view. Both of these findings are consistent with the hypothesis that leadership self-identity interacts with orientation toward leadership. More importantly, these results add to the preliminary evidence that what people think about themselves as leaders (or not) AND what they think leadership is, are two important factors that may interact in predicting important leadership outcomes.
97 CHAPTER 8 General Discussion In this thesis I have argued and provided preliminary empirical evidence that perceptions about the self within the leadership domain have tangible consequences. Following in the footsteps of substantial literature on self-views (e.g., Cross & Markus, 1994; Cyranowski & Andersen, 2000; Froming, Nasby, & McManus, 1998; Gabriel & Gardner, 1999) the present research provided the first direct assessment of leadership self-identity as a multi-dimensional construct, examining descriptive, certainty, and importance dimensions. But it is not just whether an individual thinks of him/herself as a leader (or not) that might have leadership consequences; consideration of the content of leadership beliefs was also investigated. Based on the theoretical writings of Bill Drath (2001) about views of the nature of leadership, an Orientation Toward Leadership scale was designed to assess leadership beliefs along three dimensions – dominance, developmental beliefs, and leadership as a shared/collective phenomenon. Results suggest that both leadership self-identity and orientation toward leadership predict interest in leadership development and have implications for the way supervisors and subordinates relate with one another. Further, the presence of a number of significant interactions suggests that leadership identity and the content of leadership beliefs should be examined together. Study 1 investigated the origins and correlates of leadership self-identity and orientation toward leadership in a sample of 454 undergraduate students at a large university in the northeast United States. Across all three dimensions, leadership self-identity was positively correlated with quantity, quality, and intensity of previous leadership experience, as well as core evaluations of the self, motivation to lead, and self-monitoring. Taken together, Study 1 helps to build the nomological network of leadership self-identity and helps to develop a picture of the individual with high leadership self-identity. This individual thinks positively about the self, has had rich experiences in leadership (not just quantity of experience), has a desire and duty to lead, and is effective at varying self-presentation according to the situation. Within individuals, leadership orientation was correlated with how one construes the self in relation to others and the extent to which equality is valued. The picture of the individual with stronger leadership orientation along each dimension that is painted by the correlations suggests that the individual higher in dominance orientation is more individualistic, and the individual who values shared leadership is more collectivistic. Individuals with a high development orientation endorsed the horizontal dimensions of both individualism and collectivism which emphasize equality and de-emphasize status differences. Study 2 investigated outcomes of leadership self-identity and orientation toward leadership among 231 faculty and nurse supervisors and subordinates in a university medical center. One of the main purposes of this study was to further examine the negative correlation reported by Engle and Lord (1997) that leaders who have stronger self views
98 about leadership have poorer quality relationships with subordinates. In addition to relationship quality, supervisor assessments of leadership potential of subordinates were also assessed to examine their dependence upon leadership self-views and leadership orientation. Results of hierarchical linear modeling analyses, which are better suited to handling the nonindependence of supervisors who provided ratings for multiple individuals, suggest that different leaders do evaluate their subordinates differently (in terms of both LMX and leadership potential), but leadership self-identity and orientation toward leadership of the supervisor are generally not the reason for those varying assessments. The inconsistency with Engle and Lord’s conclusion could be due to at least three factors. First, hierarchical modeling is less likely to falsely reject the null hypothesis that there is an effect. Whereas the present study used hierarchical linear modeling, Engle and Lord (1997) utilized correlational analyses. An exploratory examination of the correlations in the present study suggests that the trend is in the negative direction (though not generally significant). Thus, it could be that failing to account for non-independence biased Engle and Lord’s results. Second, it is possible that the results in either study were highly sample-dependent. Engle and Lord’s study included employees from a marketing department at an electric company, whereas the present research relied on doctors and nurses. People with different personalities tend to cluster in certain industries and organizations (Schneider, Smith, Taylor, & Fleenor, 1998) and these effects may translate into different relational styles, leadership self-identities, and orientations toward leadership across the two samples. Additional research in a variety of settings, industries, and countries might help to better understand the generalizability of the results from the present study. The third reason for the incongruous findings between the present research and Engle and Lord’s findings are the different measures of leadership self-schema. The present research relied on a direct approach that is more consistent with existing research in the selfschema literature, whereas Engle and Lord’s measure assessed organizationally endorsed leadership traits and then calculated self-schema scores by using self-reports of the ratings on those organizationally-endorsed leader traits. When self-identity for leadership and relationship quality were examined from different sources (supervisor self-identity importance, and mean subordinate ratings of LMX with their supervisor), however, there was a general negative relationship, but only in the case of the importance dimension of self-identity. In combination with the HLM results, this suggests that leadership self-identity importance of supervisors does have a tangible negative impact on relationship quality, but not from the perspective of the supervisors’ views of relationship quality (as found by Engle & Lord, 1997). Exactly how this effect may occur is not known from the data. Evidence from subordinate ratings of supervisor behavioral repertoire suggests that the negative effects of identity importance on relationships are not due to changes in behavioral repertoire. One possibility is that supervisors who have a highly important leadership self-identity may emphasize their status differential, causing poorer quality relationships. Supervisor ratings of dominance orientation, which would similarly be
99 expected to emphasize status differences, displayed a positive trend, albeit insignificant, with identity importance. Study 2 also investigated the relationship between both self-identity and leadership orientation and interest in leadership development for the combined supervisor/subordinate sample and found that the importance and descriptiveness dimensions of leadership selfidentity was moderately-to-strongly correlated with interest in leadership development (r = .58 and .53, respectively. Those who have a leadership identity, then, are more likely to be interested in leadership development, and perhaps will engage in leadership development more regularly. Finally, the belief that leadership can be developed was related to interest in leadership development, providing convergent validity evidence for the development orientation measure. Leadership Self-Identity and Orientation Toward Leadership Measures Both the Leadership Self-Identity measure and the Orientation Toward Leadership measure were explicitly designed in this study to assess two theoretically important, but as yet unmeasured, leadership constructs. A measure of leadership self-identity was designed based on previous self-schema research suggesting that individuals hold multidimensional views about themselves in a given domain (e.g., Cross & Markus, 1994; Pelham & Swann, 1989; Swann & Pelham, 2002). The extent to which leadership self-views are descriptive of oneself, as well as the certainty and importance one ascribes to those self views, help to complete the multidimensional picture of the leadership self-identity construct. The three dimensions displayed adequate reliability in Study 1 and Study 2 and were correlated with each other (indeed the descriptiveness and importance dimensions may form a single factor), but were differentially predictive of and predicted by variables in both studies. The descriptiveness and importance dimensions were most strongly predictive of/related to variables measured in both Study 1 and Study 2, whereas the certainty dimension was consistently less related to variables of interest. At this early stage, it would be prudent to examine all three self-identity dimensions in future work both separately and together to determine their relative importance in reflecting the underlying construct of leadership selfidentity. Evidence from the present research suggests that these factors should not simply be lumped into an aggregate measure, but that analysis of the component dimensions should also be considered. The Orientation Toward Leadership measure was theoretically composed of three unique factors, and items were subject to a Q-sort task before survey administration in both Study 1 and Study 2. Following administration of the Orientation Toward Leadership measure in both samples, it was refined through a process of item analysis and both exploratory and confirmatory factor analysis. Results of the confirmatory factor analyses suggests that a resulting 16-item, three factor measure fit the data well. The correlations between the three dimensions of leadership orientations (Dominance, Development, and Shared) were modest, suggesting that they tap into non-overlapping beliefs about the nature and source of leadership.
100 Relationship Between Leadership Self-Identity and Orientation Toward Leadership In both studies, there were some significant relationships between Leadership Self-Identity and Orientation Toward Leadership dimensions. However, the magnitude of these correlations was small in both samples indicating that these variables are minimally related. These low correlations, which accounted for no more than 3.2% of the total variance, are noteworthy because they suggest that seeing the self as a leader (or not) is essentially unrelated to what are one’s cognitive views of effective leadership. Having a highly selfimportant leadership identity does not mean that one is necessarily highly dominant; nor does it suggest that a strong leadership self-identity prevents one from valuing a shared approach to leadership. While both may act in concert to predict relevant leadership behaviors, they are not codependent. Limitations and Future Research Although this thesis helps to understand leadership self-identity and orientation toward leadership, there are a number of limitations to this study. Future research might help to overcome some of these limitations in addition to further exploring some of the more promising findings from this study. Both study limitations and future research possibilities are presented below. Measurement (technical and theoretical). The first limitation is a potential limitation resulting from the way that the Orientation Toward Leadership measure was developed. By requiring participants to respond to a survey that the researcher has designed, there are limits to the richness of responses that can be achieved (Hinkin, 1995). When little is known about a phenomenon, qualitative research can provide rich information. Given that little is known about “theories” of leadership beyond Drath’s work and implicit leader trait theories (e.g., Epitropaki & Martin, 2005), future research might step back to develop a measure of orientation toward leadership using rich phenomenological techniques. Rather than starting from a framework positing three different orientations, other dimensions might emerge from exploratory interviews and observation. Understanding other broad views and beliefs about leadership would help to add to the conceptual space of leadership orientation captured by the measure developed here. In addition to measurement issues associated with capturing the full construct space of leadership orientation, another way to better understand leadership orientation, as well as self-identity, may be to take an alternative approach to assessment. Non-transparent measures such as computerized reaction-time methods may be effective at providing an alternative assessment of leadership orientations and self. This technique, called Implicit Association Testing (IAT) has recently been gaining momentum in the social cognition literature (e.g., Dovidio, Kawakami, Johnson, Johnson, & Howard, 1997; Greenwald, Nosek, & Banaji, 2003; Greenwald, McGhee, & Schwartz, 1998) and may be a valuable method for obtaining responses to attitude measures that might otherwise be masked (through conscious or unconscious processes). IAT has been used to measure implicit self-esteem, and has been found to have a low degree of overlap with explicit pencil and paper methods for assessing
101 self-esteem, at the same time accounting for unique variance in lab-based outcome measures. Implicit measures of leadership self-identity do not currently exist, but could presumably be developed in the future and may provide a unique view of leadership self-identity beyond explicit measures. There remain some questions about IAT interpretation (Fazio & Olson, 2003), although some of the questions are possible to resolve (Blanton & Jaccard, in press) suggesting that this method, while imperfect, may provide a useful alternative. Another multi-method approach to assessing leadership orientation and self-identity could be simpler and involve others’ assessments, which may be particularly relevant for assessing orientation toward leadership. Subordinate assessment of supervisors’ leadership orientation may provide insight into how that supervisor leads, and was not assessed in the present study. If leadership is in the eye of the beholder (Lord & Maher, 1991), perhaps impressions of leadership orientation help to ascribe leadership or non-leadership status to an individual. Future research might also examine orientation toward leadership from both a normative and ideal perspective. A number of the items in the Orientation Toward Leadership measure asked about “effective” leadership, while others simply inquired about leadership (not stating “effective”). Without asking participants about their frame of reference and how they were interpreting the items, it is not possible to rule out that some items were interpreted as being about ideal leadership (particularly those that contained words such as “effective”, or “leaders should…”), while others were interpreted as being about normative, or typical, leadership. This conceptual blurriness may have been the cause of the inconsistent factor structures between the two samples. For example, the student sample may have been interpreting items as being about ideal leadership, while the organizational sample (where leadership examples and norms might be more salient) may have answered most items with a normative frame of reference. Future research should consider both ideal and normative perspectives separately by making two versions of the Orientation Toward Leadership measure. There may be important information in both normative and ideal measures individually, as well as important information carried in the discrepancy between normative and ideal scores. Factor structure of Orientation Toward Leadership measure. A relatively parsimonious model of the factor structure of the Orientation Toward Leadership measure was produced in the present study, although results of the exploratory factor analysis in each of the two samples suggest differences in the overall factor structure between the two samples when all 37 original items were included in the analyses. The strategy adopted in the present study was to seek only those items that loaded clearly and consistently in both samples so that a parsimonious factor structure would result. Another approach that might be considered in future research is to explicitly model the factor structure of different samples under the view that factor structures themselves might convey useful information. Perhaps different groups of people have fundamentally different approaches to understanding leadership and the ways that different leadership behaviors are related? It could be that shared leadership items may constitute several different factors in certain contexts. Or, the influence component of the proposed development/influence factor might emerge.
102 The fact that there were differences in factor loadings and factor structure between the various samples in the present study is perhaps due to differences between the samples. The student sample consisted of undergraduates from psychology and business courses; an environment that would not be characterized as leadership development intense. The doctors and nurses in Study 2, however, were part of an organization that continually promoted leadership development at all levels. The medical center even had an internal leadership development center that actively promoted courses and workshops. Although no clear patterns emerged in factor structure differences between the two samples, the goal of the present research was parsimony in measure development, which may have obscured interesting differences. Explicit modeling and comparison of the factor structure of leadership orientations in different organizations, individuals in different occupations, and different cultures might provide valuable information in and of itself. Variables assessed. Survey length constraints and difficulties surrounding access to data were particularly salient in the medical center sample. The organization’s previous experience with surveys of employees suggested that the survey used in the present research needed to be as short as possible, and a request by a key stakeholder to add a 35-item measure of leadership development climate added to the survey length. These factors contributed to the individualism and collectivism measures being dropped from the survey. In addition, the need for brevity also made it not possible to collect ratings of overall performance or actual participation in leadership development from both supervisors and subordinates. The measurement of additional variables in future research will contribute to our understanding of the importance of conceptual correlates and outcomes associated with both leadership self-identity and orientation toward leadership. Same source bias and non-response bias. For both Study 1 and Study 2, many analyses were conducted using data from the same source. In the case of Study 1, all variables were assessed by the same source, which poses a threat to the validity of the findings (Shadish, Cook, & Campbell, 2002). Non-response bias may also pose a threat to the external validity of the findings, since there may be some shared characteristics of people who did not respond as compared to those who did. Sample size. Particularly for the hierarchical analyses in Study 2, the small sample size of the supervisor group (n = 36-41, depending on the analyses in question) made it difficult to detect effects. Indeed, Kreft (1996; cited in Hofmann, Griffin, & Gavin, 2000) suggests that adequate power may not be achieved for most analyses unless there are approximately 30 groups (leaders) of 30 individuals in each group (under each leader). Thus, there may have been small effects of leadership self-identity and leadership orientation on ratings of leadership potential and LMX, but the relatively small sample size did not allow a detection of those effects.
103 Implications The findings from the present research have several implications for both research and practice of leadership and leadership development. A discussion of two general implications is presented below (implications more specific to Study 1 or Study 2 are presented in the Study 1 and Study 2 Discussion sections). In assessing developmental patterns of Army officers over time, Mumford and colleagues (Mumford, Zaccaro, Johnson, Diana, Gilbert, & Threlfall, 2000) found that differential development patterns appear to be based on both the timing and rate of growth of requisite skills. In addition to contextual factors examined as predictors of leadership development in their study, important individual-level predictors of leadership development might also be important; notably leadership self-identity and orientation toward leadership (particularly a development orientation). Leadership self-identity and development orientation, as predictors of interest in leadership development in the present study, may result in an increase in the likelihood of seeking out and obtaining the leadership skills and knowledge (i.e., actual development behaviors). Thus, over time, we might expect that leadership self-identity is related to increased leadership development. This finding is consistent with the self-schema literature, which has shown that self-schematicity leads to a seeking out of opportunities to perform and learn in a given domain (e.g., Estabrooks & Courneya, 1997; Ng, 2005). Intervening in this cyclical process may be difficult, but could involve attempting to change cognitions around the definition of leadership so that individuals might see leadership as being more possible for them (Mumford & Manley, 2003). Another possibility to explore is that intervention might occur by actively getting people involved in leadership behaviors and positions to increase self-identity and selfefficacy around leadership, even among those who do not possess a strong leadership selfidentity. Behaviors often precede cognitions and views of the self (Festinger, 1957; Cross & Markus, 1994), suggesting that a non-threatening form of behavioral change may have implications for subsequent views and, in turn, subsequent behaviors around leadership. The finding that there were considerable group-level effects is interesting and particularly noteworthy was the between-group variance in LMX ratings, which accounted for 49% of the total rating variance. This means that some supervisors have better or worse relationships with their subordinates as a whole (or at least report that they do). Originally, LMX theory focused on the different relationships that supervisors had with each subordinate, although there is now the recognition that some leaders have higher or lower quality LMX on the whole with their subordinates (Graen & Uhl-Bien, 1995) and an associated suggestion that leaders need to focus on effective relationships with all subordinates. To the extent that LMX is an important outcome because of its myriad effects such as on overall job performance, commitment, satisfaction, and turnover intentions (Gerstner & Day, 1998), as well as a key component to effective leadership development (Uhl-Bien, 2003), this finding suggests that the organization might pay attention to training leaders to have better relationships with their subordinates, as there was substantial variability accounted for at the leader level.
104 Conclusion This study contributes to the leadership literature by developing and providing preliminary validity evidence for measures of leadership self-identity and orientation toward leadership; two conceptually relevant but largely unmeasured variables. Relationships between measures of leadership self-identity, orientation toward leadership, and individual difference variables were generally significant, and in the hypothesized direction. The pattern of relationships helps to grow the nomological network of both leadership selfidentity and orientation toward leadership, and helps to provide evidence of both convergent and discriminant validity of these measures. The extent to which both leadership selfidentity and orientation toward leadership are related to outcomes of interest was more mixed, but on the whole, provide some evidence that these two newly conceptualized and measured variables are worthy of continued research and understanding.
105 REFERENCES Altemeyer, B. (1981). Right-wing authoritarianism. Manitoba: University of Manitoba Press. American Association of Medical Colleges (2003). U.S. Medical School Faculty, 2003. Available online at http://www.aamc.org/data/facultyroster/reports.htm. Andersen, B. L., Cyranowski, J. M., & Espindle, D. (1999). Men’s sexual self-schema. Journal of Personality and Social Psychology, 76, 645-661. Anderson, J. C., & Gerbing, D. W. (1991). Predicting the performance of measures in a confirmatory factor analysis with a pretest assessment of their substantive validities. Journal of Applied Psychology, 76, 732-740. Aron, A., Aron, E. N., Tudor, M., & Nelson, G. (1991). Close relationships as including other in the self. Journal of Personality and Social Psychology, 60, 241-253. Avants, S. K., Margolin, A., & Kosten, T. R. (1996). Influence of treatment readiness on outcomes of two pharmacotherapy trials for cocaine abuse among methadonemaintained patients. Psychology of Addictive Behaviors, 10, 147-156. Bandura, A. (1982). Self-efficacy mechanism in human agency. American Psychologist, 37, 122-147. Bass, B. M. (1985). Leadership and performance beyond expectations. New York: Free Press. Bass, B. M., & Avolio, B. J. (1989). Potential biases in leadership measures: How prototypes, leniency, and general satisfaction relate to ratings and rankings of transformational and transactional leadership constructs. Educational and Psychological Measurement, 49, 509-527. Bass, B. M., & Avolio, B. J. (1994). Improving organizational effectiveness through transformational leadership. Thousand Oaks, CA: Sage. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238-246. Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness of fit in the analysis of covariance structures. Psychological Bulletin, 88, 588-606. Blanton, H., & Jaccard, J. (in press). Arbitrary metrics in psychology. American Psychologist.
106 Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In Klein, K. & Kozlowski, S. W. J. (Eds.), Multilevel theory, research, and methods in organizations: Foundations extensions, and new directions (pp. 349-381). San Francisco: Jossey-Bass. Bliese, P. D., & Ployhart, R. E. (2002). Growth modeling using random coefficient models: Model building, testing, and illustrations. Organizational Research Methods, 5, 362387. Brockner, J. (1988). Self-esteem at work: Research, theory, and practice. Lexington MA: Lexington. Brockner, J., Chen, Y. R., Mannix, E. A., Leung, K., & Skarlicki, D. P. (2000). Culture and procedural fairness: When the effects of what you do depend on how you do it. Administrative Science Quarterly, 45, 138-159. Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS: Basic concepts, applications, and programming. Mahwah, NJ: Erlbaum. Cartwright, D., & Zander, A. (1960). Group dynamics research and theory. Evanston, IL: Row, Peterson. Catrambone, R., & Markus, H. (1987). The role of self-schemas in going beyond the information given. Social Cognition, 5, 349-368. Cattell, R. B. (1965). The scientific analysis of personality. Baltimore: Penguin. Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245-276. Chan, K.-Y. (2000). The relation between vocational interests and the motivation to lead. Journal of Vocational Behavior, 57, 226-245. Chan, K.-Y., & Drasgow, F. (2001). Toward a theory of individual differences and leadership: Understanding the motivation to lead. Journal of Applied Psychology, 86, 481-498. Chan, D. & Schmitt, N. (1997). Video-based versus paper-and-pencil method of assessment in situational judgment tests: Subgroup differences in test performance and face validity perceptions. Journal of Applied Psychology, 82, 143-159. Cohen, M. D., & March, J. G. (1974). Leadership and ambiguity: The American college president. New York: McGraw-Hill.
107 Cortina, J. M., Chen, G., & Dunlap, W. P. (2001). Testing interaction effects in LISREL: Examination and illustration of available procedures. Organizational Research Methods, 4, 324-360. Costa, P. T., & McCrae, R. R. 1992. Revised NEO personality inventory and NEO five-factor inventory. Odessa, FL: Psychological Assessment Resources. Crocker, J., Luhtanen, R. K., Cooper, M. L., Bouvrette, A. (2003). Contingencies of selfworth in college students: Theory and measurement. Journal of Personality and Social Psychology, 85, 894-908. Cross, S. E., Bacon, P. L., & Morris, M. L. (2000). The relational-interdependent selfconstrual and relationships. Journal of Personality and Social Psychology, 78, 791808. Cross, S. E., & Markus, H. R. (1994). Self-schemas, possible selves, and competent performance. Journal of Educational Psychology, 86, 423-438. Cyranowski, J. M., & Andersen, B. L. (2000). Evidence of self-schematic cognitive processing in women with differing sexual self-views. Journal of Social and Clinical Psychology, 19, 519-543. D’Aveni, R. (1994). Hypercompetition: Managing the dynamics of strategic maneuvering. New York: Free Press. Day, D. V., Schleicher, D. J., Unckless, A. L., & Hiller, N. J. (2002). Self-monitoring personality at work: A meta-analytic investigation of construct validity. Journal of Applied Psychology, 87, 390-401. Day, D. V., Sin, H.-P., & Chen, T. T. (2004). Assessing the burdens of leadership: Effects of formal leadership roles on individual performance over time. Personnel Psychology, 57, 573-605. DeBono, K. G., Green, S., Shair, J., & Benson, M. (1995). Attitude accessibility and biased information processing: The moderating role of self-monitoring. Motivation and Emotion, 19, 269-277. den Hartog, D. N., House, R. J., Hanges, P. J., Ruiz-Quintanilla, S. A., Dorfman, P. W., et al. (1999). Culture specific and cross-culturally generalizable implicit leadership theories: Are attributes of charismatic/transformational leadership universally endorsed? Leadership Quarterly, 10, 219-256. Denison, D. R., Hooijberg, R., & Quinn, R. E. (1995). Paradox and performance: Toward a theory of behavioral complexity in managerial leadership. Organization Science, 6, 524-540.
108 Department of Health and Human Services (2001). The registered nurse population: National sample survey of registered nurses. Available online at: ftp://ftp.hrsa.gov/bhpr/nursing/sampsurvpre.pdf. Dillman, D. A. (2000). Mail and internet surveys: The tailored design method (2nd ed.). New York: Wiley. Dovidio, J. F., Kawakami, K., Johnson, C., Johnson, B., Howard, A. (1997). On the nature of prejudice: Automatic and controlled processes. Journal of Experimental Social Psychology, 33, 510-540. Dozois, D. J., & Dobson, K. S. (2003). The structure of the self-schema in clinical depression: Differences related to episode recurrence. Cognition and Emotion, 17, 933-941. Drath, W. (2001). The deep blue sea: Rethinking the source of leadership. San Francisco, CA: Jossey-Bass. Dunning, D., & Hayes, A. F. (1996). Evidence for egocentric comparisons in social judgment. Journal of Personality and Social Psychology, 71, 213-230. Dvir, T., Eden, D., Avolio, B. J., & Shamir, B. (2002). Impact of transformational leadership on follower development and performance: A field experiment. Academy of Management Journal, 45, 735-744. Engle, E. M., & Lord, R. G. (1997). Implicit theories, self-schemas, and leader-member exchange. Academy of Management Journal, 40, 988-1010. Epitropaki, O., & Martin, R. (2004). Implicit leadership theories in applied settings: Factor structure, generalizability, and stability over time. Journal of Applied Psychology, 89, 293-310. Epitropaki, O., & Martin, R. (2005). From ideal to real: A longitudinal study of the role of implicit leadership theories on leader-member exchanges and employee outcomes. Journal of Applied Psychology, 90, 659-676. Estabrooks, P., & Courneya, K. S. (1997). Relationships among self-schema, intention, and exercise behavior. Journal of Sport and Exercise Psychology, 19, 156-168. Fazio, R. H., & Wilson, M. A. (2003). Implicit measures in social cognition research: Their meaning and use. Annual Review of Psychology, 54, 297-327. Feasel, K. E. (1995). Mediating the relation between goals and subjective well-being: global and domain-specific variants of self-efficacy. Unpublished master’s thesis. University of Illinois at Urbana-Champaign.
109 Festinger, L. (1957). A theory of cognitive dissonance. Stanford, CA: Stanford University Press. Fiedler, F. E. (1995). Cognitive resources and leadership performance. Applied Psychology: An International Review, 44, 5-28. Fiske, S. T. (1992). Thinking is for doing: Portraits of social cognition from daguerreotype to laserphoto. Journal of Personality and Social Psychology, 63, 877-889. Fiske, S. T., & Taylor, S. E. (1991). Social cognition (2nd ed.). New York: Mcgraw-Hill. Follett, M. P. (1924). Creative experience. New York: Longmans Green. Forehand, M. R., Deshpandé, R., & Reed, A. (II). 2002. Identity salience and the influence of differential activation of the social self-schema on advertising response. Journal of Applied Psychology, 87, 1086-1099. Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18, 39-50. Froming, W. J., Nasby, W., & McManus, J. (1998). Prosocial self-schemas, self-awareness, and children’s prosocial behavior. Journal of Personality and Social Psychology, 75, 766-777. Gabriel, S., & Gardner, W. L. (1999). Are there “his” and “hers” types of interdependence? The implications of gender differences in collective versus relational interdependence for affect, behavior, and cognition. Journal of Personality and Social Psychology, 77, 642-655. Gangestad, S. W., & Snyder, M. (2000). Self-monitoring: Appraisal and reappraisal. Psychological Bulletin, 126, 530-555. Gelfand, M. J., Bhawuk, D. P. S., Nishii, L. H., & Bechtold, D. J. (2004). Individualism and collectivism. In R. J. House, P. J. Hanges, M. Javidan, et al. (Eds.), Culture, leadership and organizations: The GLOBE study of 62 societies. Thousand Oaks, CA: Sage. Gergen, K. J. (1968). Personal consistency and the presentation of self. In C. Gordon & K. J. Gergen (Eds.), The self in social interaction (Vol. 1). New York: Wiley. Gerstner, C. R., & Day, D. V. (1997). Meta-analytic review of leader-member exchange theory: Correlates and construct issues. Journal of Applied Psychology, 82, 827-844.
110 Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leader-member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. Leadership Quarterly, 6, 219-247. Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring individual differences in implicit cognition: The Implicit Association Test. Journal of Personality and Social Psychology, 74, 1464-1480. Greenwald, A. G., Nosek, B. A., & Banaji, M. R. (2003). Understanding and using the Implicit Association Test: I. An improved scoring algorithm. Journal of Personality and Social Psychology, 85, 197-216. Hays, W. L. (1994). Statistics (5th ed.). Fort Worth: Harcourt Brace. Heine, S. J., & Renshaw, K. (2002). Interjudge agreement, self-enhancement, and liking: Cross-cultural divergences. Personality and Social Psychology Bulletin, 28, 578-587. Hiller, N. J., Day, D. V., & Vance, R. J. (in press). Collective Enactment of Leadership Roles and Team Effectiveness: A Field Study. The Leadership Quarterly. Hiller, N. J., & Hambrick, D. C. (2005). Conceptualizing executive hubris: The role of (hyper-) core self-evaluations in strategic decision-making. Strategic Management Journal, 26, 297-319. Hinkin, T. R. (1995). A review of scale development practices in the study of organizations. Journal of Management, 21, 967-988. Hofmann, D. A. (1997). An overview of the logic and rationale of hierarchical linear models. Journal of Management, 23, 723-744. Hofmann, D. A., Griffin, M. A., & Gavin, M. B. (2000). The application of hierarchical linear modeling to organizational research. In Klein, K. & Kozlowski, S. W. J. (Eds.), Multilevel theory, research, and methods in organizations: Foundations extensions, and new directions (pp. 349-381). San Francisco: Jossey-Bass. Hofstede, G. (1980). Culture’s consequences. Beverly Hills, CA: Sage. Hogan, J., & Hogan, R. (2002). Leadership and sociopolitical intelligence. In R. E. Riggio and S. E. Murphy (Eds.), Multiple intelligences and leadership: LEA’s organization and management series. Mahwah, NJ: Lawrence Erlbaum. Holden, R. R., & Jackson, D. N. (1979). Item subtlety and face validity in personality assessment. Journal of Consulting and Clinical Psychology, 47, 459-468. Hooijberg, R. (1996). A multidirectional approach toward leadership: An extension of the concept of behavioral complexity. Human Relations, 49, 917-946.
111 Hooijberg, R., Hunt, J. G., & Dodge, G. E. (1997). Leadership complexity and development of the leaderplex model. Journal of Management, 23, 375-408. House, R. J., & Aditya, R. N. (1997). The social scientific study of leadership: Quo Vadis? Journal of Management, 23, 409-473. Hunt, J. G., Boal, K. B., & Sorenson, R. L. (1990). Top management leadership Inside the black box. Leadership Quarterly, 1, 41-65. Jöreskog, K. G. & Sörbom, D. (1989). LISREL 7: A guide to the program and applications (2nd ed.). Chicago: SPSS. Jöreskog, K. G., & Sörbom, D. (1993). LISREL 8: Structural equation modeling with the SIMPLIS command language. Chicago: Scientific Software. Judge, T. A., Bono, J. E., Ilies, R., & Gerhardt, M. W. (2002). Personality and leadership: A qualitative and quantitative review. Journal of Applied Psychology, 87, 765-780. Judge, T. A. Erez, A., Bono, J. E., & Thoresen, C. J. (2002). Are measures of self-esteem, neuroticism, locus of control, and generalized self-efficacy indicators of a common core construct? Journal of Personality and Social Psychology, 83, 693-710. Judge, T. A. Erez, A., Bono, J. E., & Thoresen, C. J. (2003). The core self-evaluations scale: Development of a measure. Personnel Psychology, 56, 303-331. Judge, T. A., Locke, E. A., & Durham, C. C. (1997). The dispositional causes of job satisfaction: A core evaluations approach. Research in Organizational Behavior, 19, 151-188. Judge, T. A., Thoresen, C. J., Pucik, V., & Welbourne, T. M. (1999). Managerial coping with organizational change: A dispositional perspective. Journal of Applied Psychology, 84, 107-122. Kardes, F. R., Sanbonmatsu, D. M., Voss, R. T., & Fazio, R. H. (1986). Self-monitoring and attitude accessibility. Personality and Social Psychology Bulletin, 12, 468-474. Katz, D., & Kahn, R. L.(1978). The Social Psychology of Organizations (2nd ed.). New York: Wiley. Kegan, R. (1994). In over our heads: The mental demands of modern life. Cambridge, MA: Harvard University Press. Keller, T. (1999). Images of the familiar: Individual differences and implicit leadership theories. The Leadership Quarterly, 10, 589-607.
112 Kendzierski, D., & Sheffield, A. (2000). Self-schema and attributions for an exercise lapse. Basic and Applied Social Psychology, 22, 1-8. Kendzierski, D., Sheffield, A., & Morganstein, M. S. (2002). The role of self-schema in attributions for own versus other’s exercise lapse. Basic and Applied Social Psychology, 24, 251-260. Kerr, S., & Jermier, J. M. (1978). Substitutes for leadership: Their meaning and measurement. Organizational Behavior and Human Performance, 22, 395-403. Kets de Vries, M. (2001). The leadership mystique: A user’s manual for the human enterprise. Upper Saddle River, NJ: Financial Times Prentice Hall. Kihlstrom, J. F., & Klein, S. B. (1994). The self as a knowledge structure. In R. S. Wyer, Jr. & T. K. Srull (Eds.), Handbook of social cognition (2nd ed., pp. 153-208). Hillsdale, NJ: Erlbaum. Kirkpatrick, S. A., Locke, E. A. (1991). Leadership: Do traits matter? Academy of Management Executive, 5, 48-59. Komives, S., Casper, J., Longerbeam, F., Mainella, F., & Osteen, L. (2004, January). Leadership Identity Development: A grounded theory and a leadership identity development model (LID). Paper presented at the International Leadership Association, Washington. Kreft, I. G., DeLeeuw, J., Aiken, L. S. (1995). The effect of different forms of centering in hierarchical linear models. Multivariate Behavioral Research, 30, 1-21. Lee, A. Y., Aaker, J. L, & Gardner, W. L. (2000). The pleasure and pains of distinct selfconstruals: The role of interdependence in regulatory focus. Journal of Personality and Social Psychology, 78, 1122-1134. Lee, F., & Tiedens, L. Z. (2001). Is it lonely at the top?: The independence and interdependence of power holders. Research in Organizational Behavior, 23, 43-91. Lennox, R. D., & Wolfe, R. N. (1984). Revision of the Self-Monitoring Scale. Journal of Personality and Social Psychology, 46, 1349-1364. Lord, R. G., & Brown, D. J. (2004). Leadership processes and follower self-identity. Mahwah, NJ: Erlbaum. Lord, R. G., Brown, D. J., & Freiberg, S. J. (1999). Understanding the dynamics of leadership: The role of follower self-concepts in the leader/follower relationship. Organizational Behavior and Human Decision Processes, 78, 167-203.
113 Lord, R. G., De Vader, C. L., & Alliger, G. M.. (1986). A meta-analysis of the relation between personality traits and leadership perceptions: An application of validity generalization procedures. Journal of Applied Psychology, 71, 402-410 Lord, R. G., Foti, R. J., & De Vader, C. L. (1984). A test of leadership categorization theory: Internal structure, information processing, and leadership perceptions. Organizational Behavior and Human Performance, 34, 343-378. Lord, R. G., & Maher, K. J. (1991). Leadership and information processing: Linking perceptions and performance. Boston: Routledge. Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N. (1996). Effectiveness correlates of transformational and transactional leadership: A meta-analytic review of the literature. Leadership Quarterly, 7, 385-425. Markus, H. (1977). Self-schemata and processing information about the self. Journal of Personality and Social Psychology, 35, 63-78. Markus, H. Cross, S. E., & Wurf, E. (1990). The role of the self-system in competence. In R. J. Sternberg & J. Kolligian, Jr. (Eds.), Competence considered (pp. 205-226). New Haven, CT: Yale University Press. Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion and motivation. Psychological Review, 98, 224-253. Markus, H. R., & Wurf, E. (1987). The dynamic self-concept: A social psychological perspective. Annual Review of Psychology, 38, 299-337. McCauley, C. D., Moxley, R. S., & Van Velsor, E. (Eds.) (1998). The center for creative leadership handbook of leadership development. San Francisco: Jossey-Bass. McCauley, C. D., Ruderman, M. N., Ohlott, P. J., & Morrow, J. E. (1994). Assessing the developmental components of managerial jobs. Journal of Applied Psychology, 79, 544-560. Merton, R.K. (1957). Social theory and social structure, NY: Free Press Mueller, J. H., Ross, M. J., & Heesacker, M. (1984). Distinguishing me from three. Bulletin of the Psychonomic Society, 22, 79-82. Mumford, M. D., & Manley, G. M. (2003). Putting the development in leadership development: Implications for theory and practice. In S. E. Murphy and R. E. Riggio (Eds.), The future of leadership development. Mahwah, NJ: Erlbaum.
114 Mumford, M. D., Zaccaro, S. J., Johnson, J. F., Diana, M., Gilbert, J. A., & Threlfall, K. V. (2000). Patterns of leader characteristics: Implications for performance and development. The Leadership Quarterly, 11, 115-134. Morrison, M. A., & Morrison, T. G. (2002). Development and validation of a scale measuring modern prejudice toward gay men and lesbian women. Journal of Homosexuality, 43, 15-37. Ng, C.-H. (2005). Modelling the relationship between self-schemas, learning, and learning outcomes in the domain of high school mathematics. Psychologia: An International Journal of Psychology in the Orient, 48, 14-30. Offerman, L. R., Kennedy, J. K., Jr., & Wirtz, P. W. (1994). Implicit leadership theories: Content, structure, and generalizability. Leadership Quarterly, 5, 43-58. Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism and collectivism: Evaluation of theoretical assumptions and meta-analyses. Psychological Bulletin, 128, 3-72. Pearce, C. L., & Conger, J. A. (Eds.). (2003). Shared leadership: Reframing the hows and whys of leadership. Thousand Oaks, CA: Sage. Pelham, B. W., & Swann, W. B., Jr. (1989). From self-conceptions to self-worth: On the sources and structure of global self-esteem. Journal of Personality and Social Psychology, 57, 672-680. Penland, E. A., Masten, W. G., Zelhart, P., Fournet, G. P., & Callahan, T. A. (2000). Possible selves, depression and coping skills in university students. Personality and Individual Differences, 29, 963-969. Petersen, L. E., Stahlberg, D., & Dauenheimer, D. (2000). Effects of self-schema elaboration on affective and cognitive reactions to self-relevant information. Genetic, Social, and General Psychology Monographs, 126, 25-42. Peterson, R. S., Smith, D. B., Martorana, P. V., Owens, P. D. (2003). The impact of chief executive officer personality on top management team dynamics: One mechanism by which leadership affects organizational performance. Journal of Applied Psychology, 88, 795-808. Pincus, T., & Morley, S. (2001). Cognitive-processing bias in chronic pain: A review and integration. Psychological Bulletin, 127, 599-617. Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dominance orientation: A personality variable predicting social and political attitudes. Journal of Personality and Social Psychology, 67, 741-763.
115 Quinn, R. E. (1988). Beyond rational management: Mastering the paradoxes and competing demands of high performance. San Francisco: Jossey-Bass. Quiñones, M. A., Ford, J. K., & Teachout, M. S. (1995). The relationship between work experience and job performance: A conceptual and meta-analytic review. Personnel Psychology, 48, 887-910. Raelin, J. A. (2004). Creating leaderful organizations: How to bring out leadership in everyone. Williston, VT: Berrett-Koehler. Reissing, E. D., Binik, Y. M., Khalifé, S., Cohen, D., & Amsel, R. (2003). Etiological correlates of vaginismus: Sexual and physical abuse, sexual knowledge, sexual selfschema, and relationship adjustment. Journal of Sex and Marital Therapy, 29, 47-59. Rokeach, M. (1972). Beliefs, attitudes, and values: A theory of organization and change. San Francisco: Jossey-Bass. Schneider, B., Smith, D. B., Taylor, S., & Fleenor, J. (1998). Personality and organizations: A test of the homogeneity of personality hypothesis. Journal of Applied Psychology, 462-470. Schyns, B., & Meindl, J. R. (2005). Implicit leadership theories: Essays and explorations. Greenwich, CT: Infoage. Singelis, T. M. (1994). The measurement of independent and interdependent self-construals. Personality and Social Psychology Bulletin, 20, 580-591. Singelis, T. M., Triandis, H. C., Bhawuk, D. P., & Gelfand, M. J. (1995). Horizontal and vertical dimensions of individualism and collectivism: A theoretical and measurement refinement. Cross-Cultural Research, 29, 240-275. Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and QuasiExperimental Designs for Generalized Causal Inference. Boston: Houghton-Mifflin. Slife, B. D., & Williams, R. N. (1995). What’s behind the research? Discovering hidden assumptions in the behavioral sciences. Thousand Oaks, CA: Sage. Smith, J. A., & Foti, R. J. (1998). A pattern approach to the study of leader emergence. The Leadership Quarterly, 9, 147-160. Snyder, M. (1974). Self monitoring of expressive behavior. Journal of Personality and Social Psychology, 30, 526-537. Snyder, M., & Cantor, N. (1980). Thinking about ourselves and other: Self-monitoring and social knowledge. Journal of Personality and Social Psychology, 39, 222-234.
116 Srull, T. K., & Wyer, R. S. Jr. (1993). The mental representation of trait and autobiographical knowledge about the self. Hillsdale, NJ: Erlbaum. Steiger, J. H. (1990). Structural model evaluation and modification: An interval estimation method. Multivariate Behavioral Research, 25, 173-180. Stogdill, R. M. (1948). Personal factors associated with leadership: A survey of the literature. Journal of Psychology, 25, 35-71. Stogdill, R. M . (1974). Handbook of leadership: A survey of theory and research. New York: Free Press. Streufert, S., & Nogami, G. (1989). Cognitive style and complexity: Implications for I/O Psychology. In C. L. Cooper and I. Robertson (Eds.), International review of industrial and organizational psychology. New York: Wiley. Stryker, S. (1980). Symbolic interactionism: A social structural version. Menlo Park: CA: Benjamin/Cummings. Swann, W. B., Jr., & Ely, R. J. (1984). A battle of wills: Self-verification versus behavioral confirmation. Journal of Personality and Social Psychology, 46, 1287-1302. Swann, W. B., Jr., & Pelham, B. W. (2002). Who wants out when the going gets good? Psychological investment and preference for self-verifying college roommates. Journal of Self and Identity, 1, 219-233. Tarquinio, C., & Somat, A. (2001). Scholastic achievement, academic self-schemata and normative clearsightedness. European Journal of Psychology and Education, 16, 117129. Tesluk, P. E., & Jacobs, R. R. (1998). Toward an integrated model of work experience. Personnel Psychology, 51, 321-355. Triandis, H. C. (1989). The self and social behavior in differing cultural contexts. Psychological Review, 96, 506-520. Triandis, H. C. (2001). Individualism-collectivism and personality. Journal of Personality, 69, 907-924. Triandis, H. C., & Gelfand, M. J. (1998). Converging measurement of horizontal and vertical individualism and collectivism. Journal of Personality and Social Psychology, 74, 118-128. Uhl-Bien, M. (2003). Relationship development as a key ingredient for leadership development. In S. E. Murphy and R. E. Riggio (Eds.), The future of leadership development. Mahwah, NJ: Erlbaum.
117 Van Velsor, E., & McCauley, C. D. (2004). Our view of leadership development. In C. D. McCauley & E. Van Velsor (Eds.), The Center for Creative Leadership handbook of leadership development (pp. 1-22). San Francisco: Jossey-Bass. Wagner, J. A. (2002). Utilitarian and ontological variation in individualism-collectivism. Research in Organizational Behavior, 24, 301-345. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54, 1063-1070. Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood. Psychological Bulletin, 98, 219-235. Wiers, R. W., de Jong, P. J., Havermans, R., & Jelicic, M. (2004). How to change implicit drug use-related cognitions in prevention: A transdisciplinary integration of findings from experimental psychopathology, social cognition, memory, and experimental learning psychology. Substance Use and Misuse, 39, 1625-1684. Yukl, G., & Lepsinger, R. (1990). Preliminary report on the validation of the Managerial Practices Survey. In K. E. Clark and M. B. Clark (Eds.), Measures of leadership (pp. 223-237). West Orange: NJ: Leadership Library of America. Yurek, D., Farrar, W., & Andersen, B. L. (2000). Breast cancer surgery: Comparing surgical groups and determining individual differences in postoperative sexuality and body change stress. Journal of Consulting and Clinical Psychology, 68, 697-709. Zaccaro, S. J., & Klimoski, R. J (2001). The nature of organizational leadership: An introduction. In S. J. Zaccaro and R. J. Klimoski (Eds.), The nature of organizational leadership. San Francisco: Jossey-Bass.
118
Appendix A Orientation Toward Leadership: The 50 initial items developed
119 Personal Dominance Core Concepts: taking charge; making things happen; maintaining control; responsible for group performance; leadership is about leader traits or qualities that leaders “possess” (fundamentally different from followers); use of position power. 1. 2. 3. 4. 5. 6. 7. 8. 9.
Leadership often involves telling other people what to do. One’s formal position in an organization determines whether or not they are a leader. Leadership and power are pretty much the same thing. A leader never shows his/her weaknesses. Leadership is about directing other people. Leadership is about controlling others (or leadership is about control). In being an effective leader, it is often necessary to use rewards. In being an effective leader, it is often necessary to use punishment. Effective leaders should try to minimize the amount that their subordinates question them. 10. Leadership is mainly about maintaining control. 11. Good leaders never order other people around (reverse). 12. You can’t teach leadership; you’re either a leader, or you’re not. 13. Leaders, as people, are fundamentally different from non-leaders. 14. When a group performs well, it is primarily the result of an effective leader. 15. If you supervise others, you are a leader. 16. Leaders have to order other people around. 17. Leadership is about taking charge of a situation 18. Being an effective leader requires using both rewards and punishment. Interpersonal Influence Core Concepts: leadership is about influence; people possess (OR CAN ACQUIRE) certain qualities and characteristics that enable them to be effective in a leadership role; leaders influence followers more than followers influence leaders. 19. Leadership is about convincing other people what to do. 20. Leadership is about using influence, rather than force. 21. Leadership is a two-way street between leaders and followers. 22. Getting “buy-in” from followers is an important leadership task. 23. People can be taught to be more effective leaders. 24. Leaders can acquire skills to make them more effective. 25. Leaders allow followers to influence them. 26. Effective leaders influence followers more than followers influence them. 27. Leadership is a role occupied by the most influential person in a group. 28. Skills and abilities for leadership can be developed. 29. Effective leaders rely on their charisma. 30. Leadership involves persuading other people. 31. Effective leaders seek different perspectives.
120
Relational Dialogue – Meaning Making – Shared Leadership Core Concepts: leadership is the property of a social system, individual people do not possess leadership; leadership happens when people participate in collaborative forms of thought and action; if there is an individual leader, the actions that person takes are an aspect of participation in the process of leadership. 32. Effective leadership requires seeking different perspectives when solving problems. 33. Leadership is about the group, rather than a single leader. 34. Leadership involves the group collectively making decisions. 35. Team members should share in leadership tasks. 36. Facing challenges, setting direction, and maintaining commitment are tasks that require the whole group. 37. Leadership is not possessed by any one individual. 38. Team members should share in planning and organizing. 39. Team members should share in problem solving. 40. Team members should share in supporting others. 41. Team members should share in development and mentoring. 42. Leadership is the responsibility of everybody in a group. 43. Individual people do not possess leadership; it is a property of the group. 44. Leadership happens when people collaborate on shared work. 45. Leadership is a group process, not the property of a single individual. 46. Leadership requires that there is a consensus. 47. Leadership can be shared. 48. All group members should participate in leadership. 49. Group members should work together in performing leadership tasks. 50. Together, group members create leadership.
121
Appendix B Q-Sort task: Instructions given to raters, description of dimensions, and items assigned to their intended construct
122 Ways of Thinking about Leadership The items below are intended to provide an adequate assessment of 3 different ways of thinking about leadership. After reading the definition and description of the three dimensions, please match each item with the appropriate dimension by typing in the corresponding number next to each item. If an item does not appear to fit any dimension, type “ND” in the space provided. Feel free to consult the dimension descriptions as often as is necessary. Once you are finished, please return the completed Q-Sort to me via email attachment. Thanks for your help! and if you have any specific comments or thoughts about the items as you complete the task, I would appreciate hearing them.
Dimension Descriptions 1
Personal Dominance- Leadership is viewed as something that a person possesses as a result of their position or individual traits (or both). In this view, leaders influence followers in a unidirectional fashion, from leader to follower. Leaders are born (and not made), and are fundamentally different people from followers. The leader is responsible for group performance. Control, power, and “taking charge” are often common themes. The core of this approach is that a defined leader is the source of leadership and followers are receivers of leadership; there is no meaningful two-way interaction.
2 Interpersonal Influence- Leadership is viewed as a negotiated process of influence, rather than simply being about power. In this leadership worldview, leadership is not simply authoritative, top-down leadership; leaders must take into account other perspectives. In cases where a formal leader in a group is not predetermined, the most influential person will generally emerge as the leader. In cases where there is a defined position of “leader”, this view presumes that leadership is more a process of influence, rather than coercion and power utilization. People possess (OR CAN ACQUIRE) certain qualities and characteristics that enable them to be effective in a leadership role. Leaders influence followers more than followers influence leaders. 3
Relational Dialogue- Leadership is viewed a shared process of meaning-making. This orientation is based on the idea that the difficult demands of modern life are too much for a single leader to handle, and the focus is on the process of leadership, which inherently includes “follower” involvement. Leadership is seen as a property of the group, rather than one individual, and is understood to emerge as a result of collaborative interaction. Leadership is seen as the shared responsibility of all group members. If there is an individual in the position of “leader”, the actions that person takes are an aspect of participation in the process of leadership.
1. __1__ Telling other people what to do is a big part of leadership. 2. __3__ Group members can work together in performing leadership tasks. 3. __1__ Leaders order other people around. 4. __2__ Leadership involves persuading other people. 5. __2__ Effective leaders seek different perspectives. 6. __3__ Individual people do not possess leadership -- it is a property of the group. 7. __1__ Leaders should minimize the amount that others question them.
123 8. __3__ Leadership happens when people collaborate. 9. __2__ The leader should be the most influential person in a group. 10. __3__ Leadership is a process. 11. __1__ Leadership and power are pretty much the same thing. 12. __3__ All group members should participate in leadership. 13. __1__ A leader never shows weakness. 14. __1__ Leadership is about taking charge of a situation. 15. __2__ People can be taught to be more effective leaders. 16. __1__ Being a leader requires using both rewards and punishment. 17. __2__ Skills and abilities for leadership can be developed. 18. __3__ Leadership is a property of the group, not the individual. 19. __1__ Leadership is about maintaining control. 20. __3__ Leadership involves a group collectively making decisions. 21. __2__ Getting “buy-in” from followers is an important leadership task. 22. __2__ Effective leaders influence followers more than followers influence them. 23. __1__ Leaders, as people, are fundamentally different from non-leaders. 24. __3__ Leadership can be shared. 25. __3__ Leadership is the responsibility of everybody in a group. 26. __1__ When a group performs well, it is largely because of the leader. 27. __2__ Leadership is about using influence. 28. __1__ One’s formal position determines whether they are a leader. 29. __2__ Leadership is a two-way street between leaders and followers. 30. __2__ Leaders can acquire skills to make them more effective. 31. __3__ Together, group members create leadership. 32. __1__ You’re either a leader, or you’re not. 33. __2__ Leaders allow followers to influence them. 34. __1__ You can’t teach leadership. 35. __3__ Leadership is about the group, rather than a single leader. 36. __3__ Leadership is not possessed by any one individual. 37. __1__ If you supervise others, you are a leader.
124
Appendix C Item-Level Descriptive Statistics for Study 1
125
Item description
N
Implicit Leadership 1 (understanding) Implicit Leadership 2: dominant Implicit Leadership 3: helpful Implicit Leadership 4:selfish Implicit Leadership 5: dynamic Implicit Leadership 6: conceited Implicit Leadership 7: funny Implicit Leadership 8: intelligent Implicit Leadership 9: manipulative Implicit Leadership 10: loud Implicit Leadership 11: knowledgeable Implicit Leadership 12: masculine Implicit Leadership 13: clever Implicit Leadership 14: educated Implicit Leadership 15: dedicated Implicit Leadership 16: sincere Implicit Leadership 17: motivated Implicit Leadership 18: hard-working Implicit Leadership 19: energetic Implicit Leadership 20: male Implicit Leadership 21: cheerful Implicit Leadership 22: strong Implicit Leadership 23: pushy Self-assessment1: understanding Self-assessment2: dominant Self-assessment3: helpful Self-Assessment4: selfish Self-Assessment5: dynamic Self-Assessment6: conceited Self-Assessment7: funny Self-Assessment8: intelligent Self-Assessment9: manipulative Self-Assessment10: loud Self-Assessment11: knowledgeable Self-Assessment12: masculine Self-Assessment13: clever Self-Assessment14: educated
454 454 454 454 454 454 453 454 454 453 454 453 454 453 454 454 454 449 453 454 453 453 453 456 456 456 454 455 456 456 455 456 456 456 456 456 456
Min Max 2 1 2 1 2 1 1 2 1 1 2 1 1 1 2 2 2 2 2 1 1 1 1 2 1 2 1 1 1 2 2 1 1 2 1 1 3
5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5
Mean
S.D.
4.12 3.12 4.20 2.01 4.11 2.11 2.97 4.26 2.65 3.02 4.37 2.48 3.93 4.13 4.61 3.95 4.61 4.56 4.14 2.26 3.30 3.62 2.50 4.20 2.56 4.20 2.30 3.64 2.01 3.94 4.16 2.32 2.92 4.02 2.66 3.83 4.28
0.75 0.94 0.75 0.89 0.77 0.94 0.88 0.67 1.06 0.91 0.67 1.06 0.80 0.74 0.59 0.87 0.61 0.62 0.77 1.14 0.88 0.94 0.91 0.63 0.96 0.69 0.84 0.87 0.97 0.77 0.63 1.06 1.14 0.63 1.44 0.78 0.62
126 Self-Assessment15: dedicated Self-Assessment16: sincere Self-Assessment17: motivated Self-Assessment18: hardworking Self-Assessment19: energetic Self-Assessment20: male Self-Assessment21: cheerful Self-Assessment22: strong Self-Assessment23: pushy CSE1 (confident get success deserve) CSE2R (feel depressed) CSE3 (generally succeed) CSE4R (when fail feel worthless) CSE5 (complete successfully) CSE6R (no control of work) CSE7 (satisfied with work) CSE8R (doubts about competence) CSE9 (determine what happens) CSE10R (no control) CSE11 (capable of coping problems) CSE12R (look pretty bleak Collectivism1 (hate to disagree) Collectivism2 (do my own thing) Collectivism3 (direct and forthright) Collectivism4 (other wellbeing important) Collectivism5 (tense when other better) Collectivism6 (succeed because of abilities) Collectivism7 (competition good) Collectivism8 (duty before pleasure) Collectivism9 (obey family) Collectivism10 (share w neighbors) Collectivism11 (enjoy unique) Collectivism12 (happiness depends on others) Collectivism13 (winning everything) Collectivism14 (impt. to do better job than others) Collectivism15 (annoy when others better) Collectivism16 (enjoy competition) Collectivism17 (no emphasize winning) Collectivism18 (proud of prized colleague) Collectivism19 (help relative with finance) Collectivism20 (important for harmony)
455 455 455 456 455 455 455 454 454 456 455 456 456 454 456 455 455 454 456 455 456 455 455 455 455 455 455 454 455 455 454 454 455 455 453 454 453 453 453 454 453
2 2 1 1 2 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 2 1 1 1 1 1 1 1 2 2
5 5 5 5 5 5 5 5 5 7 7 7 7 6 7 7 7 7 7 7 7 6 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
4.21 4.35 4.04 4.21 4.01 2.91 3.86 3.77 2.16 4.73 3.42 4.73 3.77 5.03 3.89 4.48 4.07 4.38 4.34 4.65 3.87 3.60 4.77 5.26 5.51 4.11 5.59 4.62 4.11 3.90 4.61 5.48 4.79 3.42 4.61 4.16 4.89 3.88 4.89 5.88 5.65
0.78 0.69 0.89 0.82 0.87 1.90 0.79 0.86 0.85 1.61 1.50 1.83 1.59 1.46 1.42 1.87 1.78 1.83 1.76 1.77 1.70 1.43 1.28 1.26 1.03 1.40 0.97 1.52 1.41 1.52 1.34 1.10 1.32 1.61 1.38 1.40 1.31 1.54 1.14 1.03 0.96
127 Collectivism21 (feel good cooperating) Collectivism22 (sacrifice activity for family) Collectivism23 (sacrifice self-interest for group) Collectivism24 (I am unique) Collectivism25 (competition is natural) Collectivism26 (child. honored parent award) Collectivism27 (pleasure time with others) Ldr experience1 how often chosen leader Ldr experience2 (take on leadership) Ldr experience3 (often others look to you) Ldr experience4 (often involved, form. or not) Ldr fdbk1 (significant amount of feedback) Ldr fdbk2 (info to receive strengths and weak) Ldr fdbk3 (useful advice from higher up) Ldr fdbk4 (tell how effective I was) Ldr fdbk5 (task-feedback) Ldr fdbk6 (feedback from peers) Ldr fdbk7 (feedback helped me lead more effect'ly) Ldr devl1 (support as I developed) Ldr devl2 (at least one mentor) Ldr devl3 (resources necessary to be better) Ldr devl4 (support when I made mistake) Ldr devl5 (allowed to develop abilities and pot) Ldr devl6 (encouraged to develop) Ldr chlg1 (had challenging experiences) Ldr chlg2 (one exp. push me to limit) Ldr chlg3 (one exp. reaching diff. goal) Ldr chlg4 (juggle competing demands) Ldr chlg5 (felt like giving up at some point) Ldr chlg6 (wide variety of leadership experiences) Ldrship Orientation1 (telling others what do) Ldrship Orientation2 (work together in leadership) Ldrship Orientation3 (order others around) Ldrship Orientation4 (persuading others) Ldrship Orientation5 (seek diff. perspective) Ldrship Orientation6 (property of group) Ldrship Orientation7 (not openly question) Ldrship Orientation8 (when people collaborate) Ldrship Orientation9 (should be most influential) Ldrship Orientation10 (ldrship is a process) Ldrship Orientation11 (ldrship = power)
456 455 454 454 455 455 453 454 453 449 454 455 456 456 456 453 456 456 454 453 449 453 454 453 454 455 453 455 453 454 455 453 454 454 454 453 455 455 449 453 449
3 1 1 1 1 2 2 1 1 1 2 1 1 1 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 1 2 1
7 7 7 7 7 7 7 5 5 5 5 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
5.71 3.92 4.12 5.70 4.78 5.86 5.66 3.41 3.64 3.52 3.65 4.67 4.93 5.33 5.29 5.23 4.98 5.07 5.19 5.61 5.36 5.49 5.56 5.57 5.32 5.08 5.54 5.57 4.24 4.87 4.66 5.85 3.48 4.90 6.04 3.49 2.55 4.90 4.91 5.50 3.33
0.90 1.47 1.26 1.07 1.46 1.12 1.11 0.82 0.86 0.79 0.84 1.33 1.30 1.24 1.11 1.02 1.30 1.24 1.32 1.32 1.19 1.11 1.03 1.15 1.40 1.52 1.18 1.23 1.74 1.48 1.30 0.98 1.40 1.28 1.01 1.50 1.43 1.10 1.35 1.09 1.52
128 Ldrship Orientation12 (everyone participate) Ldrship Orientation13 (never shows weakness) Ldrship Orientation14 (is about taking charge) Ldrship Orientation15 (can teach more effective) Ldrship Orientation16 (use rewards and punish) Ldrship Orientation17 (skill can develop) Ldrship Orientation18 (property of group) Ldrship Orientation19 (maintaining control) Ldrship Orientation20 (group collective decisions) Ldrship Orientation21 (buy-in is important) Ldrship Orientation22 (leaders influence more) Ldrship Orientation23 (fundamentally different) Ldrship Orientation24 (can be shared) Ldrship Orientation25 (responsibility of everyone) Ldrship Orientation26 (leader causes performance.) Ldrship Orientation27 (= influence) Ldrship Orientation28 (formal pos. = ldrship) Ldrship Orientation29 (two-way street) Ldrship Orientation30 (can acquire skills) Ldrship Orientation31 (group create ldrship) Ldrship Orientation32 (either leader or not) Ldrship Orientation33 (leaders allow follower infl.) Ldrship Orientation34 (can't teach ldrship) Ldrship Orientation35 (ldrship about group) Ldrship Orientation36 (ldrship not poss. by one) Ldrship Orientation37 (supervisor = leader) self-identity descriptive1 (I am a leader) self-id. descriptive2 (I see myself as leader) self-id. descriptive3 (describe me to others) self-id. descriptive4 (prefer seen by others) selfcertainty1 (how certain) selfcertainty2 (how certain) selfcertainty3 (how certain) selfcertainty4 (how certain) self-importance1 (how important) self-importance2 (how important) self-importance3 (how important) self-importance4 (how important) ldrcon1 (confident ldr in most situations) ldrcon2 (confident lead effectively) ldrcon3 (confident lead group effectively)
449 454 454 455 453 454 454 454 453 449 454 455 453 454 454 453 454 453 454 454 455 454 454 455 454 455 454 453 454 454 453 449 453 453 454 454 455 455 455 454 455
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
4.77 3.75 5.43 5.74 5.23 5.68 4.04 5.00 4.73 4.79 4.70 4.28 5.55 4.48 4.40 4.87 3.42 5.38 5.83 4.73 3.68 4.73 3.13 4.55 3.69 3.70 5.08 4.55 4.93 5.03 5.28 5.37 5.19 5.33 4.72 4.68 4.23 4.79 5.64 5.75 5.82
1.29 1.64 1.05 1.11 1.15 1.03 1.40 1.16 1.18 1.24 1.41 1.44 1.03 1.25 1.36 1.22 1.43 1.19 0.99 1.23 1.57 1.28 1.48 1.35 1.41 1.50 1.43 1.67 1.59 1.53 1.44 1.48 1.58 1.52 1.52 1.59 1.57 1.61 1.08 1.02 0.96
129 ldrcon4 (confident ability to influence group) ldrcon5 (confident help overcome obstacles) ldrcon6 (confident ability to set direction) ldrcon7 (confident ability to gain commitment) mtlead1 (prefer being leader) mtlead2r (never agree ld just because vote) mtlead3 (duty to lead if asked) mtlead4 (lead even if no special rewards) mtlead5 (like being in charge of others) mtlead6 (never expect benefits when leader) mtlead7 (usually want to be leader) mtlead8 (not right to decline leadership roles) mtlead9 (tendency to take charge) mtlead10r (leading is dirty job more than honor) mtlead11r (only interested lead if advantages) mtlead12 (agree to lead whenever asked) mtlead13 (never agree unless benefits) mtlead13 (never agree unless benefits) mtlead14r (actively support but prefers not to lead) mtlead15 (honor and privilege to be asked) mtlead16r (only lead if benefit) mtlead17r (what's in it for me) mtlead18r (not interested to lead others) mtlead19r (enough of my own problems to worry) mtlead20 (taught value in leading others) mtlead21 (seldom reluctant to be leader) mtlead22 (appropriate to accept when asked) mtlead23r (I contribute more to group if follower) mtlead24 (taught I should always volunteer to ld) mtlead25r (definitely not a leader by nature) mtlead26 (never expect more privileges if leader) mtlead27 (people should volunteer over waiting) selfmonitoring1 (ability to alter behavior if necc.) selfmonitoring2 (often able to read true emotions) selfmonitoring3 (ability control how come across) selfmonitoring4 (sensitive to slightest facial change) selfmonitoring5 (good intuition for others motives) selfmonitoring6 (can tell when joke bad taste) selfmonitoring7 (readily change image when necc.) selfmonitoring8 (tell if I said something inapprop.) selfmonitoring9 (trouble changing behavior)
455 454 454 455 454 455 456 453 454 456 455 456 453 455 456 453 455 455 453 455 454 453 454 455 455 455 454 456 453 454 456 453 456 455 455 456 456 455 456 455 455
2 2 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 1 1
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
5.56 5.83 5.82 5.68 5.12 4.65 5.32 5.70 4.56 4.85 4.78 3.55 4.83 5.16 5.42 4.98 2.63 5.37 4.38 5.68 5.23 5.22 5.11 5.13 4.92 4.48 4.76 4.62 4.24 4.94 4.55 4.64 5.51 5.26 5.28 5.03 5.44 5.44 5.08 5.52 5.04
1.09 0.97 1.02 1.06 1.39 1.43 1.14 0.98 1.38 1.29 1.35 1.37 1.42 1.46 1.33 1.29 1.34 1.34 1.54 1.02 1.40 1.45 1.50 1.39 1.24 1.47 1.22 1.39 1.34 1.51 1.46 1.20 1.07 1.21 1.22 1.28 1.21 1.04 1.16 1.00 1.29
130 selfmonitoring10 (adjust to meet req. of sit) selfmonitoring11 (can tell if someone lying) selfmonitoring12r (difficulty putting up good front) selfmonitoring13 (easy to regulate actions)
453 455 453 456
1 1 1 1
7 7 7 7
5.22 4.97 4.71 5.41
1.14 1.24 1.46 1.01
131
Appendix D Item Level Descriptive Statistics for Study 2 (Medical Center Sample)
132
Item description Ldrship Orientation 1 (telling others what to do) Ldrship Orientation 2 (work together in ldrshp tasks) Ldrship Orientation3 (order others around) Ldrship Orientation4 (persuading others) Ldrship Orientation5 (seek diff. perspectives) Ldrship Orientation6 (property of a group) Ldrship Orientation7 (minimize questioning) Ldrship Orientation8 (when people collaborate) Ldrship Orientation9 (most influential in group) Ldrship Orientation10 (is a process) Ldrship Orientation11 (ldrship = power) Ldrship Orientation12 (everyone participate) Ldrship Orientation13 (never shows weakness) Ldrship Orientation14 (= taking charge) Ldrship Orientation15 (can be taught more effective) Ldrship Orientation16 (use rewards and punish) Ldrship Orientation17 (skills can develop) Ldrship Orientation18 (property of group) Ldrship Orientation19 (=maintaining control) Ldrship Orientation20 (group collective decisions) Ldrship Orientation21 (buy-in is important) Ldrship Orientation22 (leaders influence more) Ldrship Orientation23 (fundamentally different) Ldrship Orientation24 (can be shared) Ldrship Orientation25 (responsibility of everyone) Ldrship Orientation26 (leader causes performance.) Ldrship Orientation27 (about use influence) Ldrship Orientation28 (formal pos. = ldrship) Ldrship Orientation29 (two way street) Ldrship Orientation30 (can acquire skills) Ldrship Orientation31 (group create leadership) Ldrship Orientation32 (either leaders or not) Ldrship Orientation33 (leaders allow follower infl.) Ldrship Orientation34 (can't be taught) Ldrship Orientation35 (ldrship about group) Ldrship Orientation36 (not possessed by any one) Ldrship Orientation37 (supervisor = leader) LSI: selfdesc1 (I am a leader)
N Min 231 1 230 2 229 1 230 1 229 2 230 1 227 1 228 1 229 1 231 1 230 1 229 1 231 1 228 1 231 2 228 1 231 2 230 1 231 1 231 1 228 1 228 1 230 1 231 2 231 1 231 1 231 1 230 1 231 2 231 2 230 2 231 1 231 1 231 1 230 1 227 1 231 1 226 1
Max 7 7 6 7 7 6 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7
Mean 3.65 5.83 2.22 4.82 6.45 2.75 2.28 5.24 3.82 5.77 2.45 5.02 2.32 5.10 6.23 4.96 6.02 4.38 4.17 5.13 5.69 4.45 3.68 6.16 4.80 3.61 4.62 2.27 5.94 6.39 5.24 3.81 5.06 5.20 5.00 4.08 2.99 4.57
S.D. 1.63 0.84 1.26 1.42 0.78 1.28 1.21 1.32 1.48 1.06 1.46 1.50 1.02 1.27 0.76 1.45 0.86 1.59 1.63 1.34 1.45 1.52 1.58 0.79 1.51 1.42 1.34 1.34 0.86 0.81 1.21 1.50 1.38 1.43 1.47 1.63 1.58 1.48
133
LSI: selfdesc2 (I see myself as a leader) LSI: selfdesc3 (describe me to others as ldr) LSI: selfdesc4 (prefer seen by others as ldr) LSI: selfcertainty1 (how certain of ratings) LSI: selfcertainty2 (how certain) LSI: selfcertainty3 (how certain) LSI: selfcertainty4 (how certain) LSI: selfimportance1 (how important to self) LSI: selfimportance2 (how important to self) LSI: selfimportance3 (how important to self) LSI: selfimportance4 (how important to self) lddev1 (if given opportunity, I'd participate) lddev2 (strong desire to develop) lddev3 (I cont. look for ways to dev. ldrship) Age (1=20-30,2=31-40,3=41-50,4=51-60,5=60+) Tenure in current position LMX item 1 mean (sup & sub) LMX item2 LMX item3 LMX item4 LMX item5 LMX item6 LMX item7 Leadership potential of Subordinates1 Leadership potential of Subordinates2 Leadership potential of Subordinates3 Leadership potential of Subordinates4 Behavioral repertoire1 (sensitive, caring way) Behavioral repertoire2 (encourages part. d.m.) Behavioral repertoire3 (shows empathy) Behavioral repertoire4 (work participatively) Behavioral repertoire5 (experiments new concepts) Behavioral repertoire6 (inventive ideas) Behavioral repertoire7 (influences decisions) Behavioral repertoire8 (exerts upward influence) Behavioral repertoire9 (foresees workflow problems) Behavioral repertoire10 (brings order to unit) Behavioral repertoire11 (tight logistical control) Behavioral repertoire12 (detects discrepancies) Behavioral repertoire13 (sees unit deliver goals)
224 225 224 226 224 225 224 226 224 225 224 227 227 227 221 222 227 226 223 227 222 227 220 44 44 44 44 183 183 183 180 182 183 175 178 180 182 179 171 181
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0.5 1 1 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 1 1 1 1
7 7 7 7 7 7 7 7 7 7 7 7 7 7 5 34 5 5 5 5 5 5 5 10 10 10 10 7 7 7 7 7 7 7 7 7 7 7 7 7
4.62 3.77 4.30 5.28 5.34 5.35 5.35 4.32 4.35 3.74 4.15 5.31 5.09 5.00 3.10 9.29 3.47 3.59 3.58 3.86 3.24 3.76 3.64 6.48 6.43 7.25 6.11 5.06 4.81 4.91 4.57 4.63 4.33 4.29 4.34 4.23 4.47 4.32 4.49 5.01
1.50 1.65 1.64 1.44 1.43 1.44 1.44 1.62 1.71 1.68 1.69 1.65 1.69 1.69 1.05 8.41 1.26 1.17 1.13 1.07 1.29 1.01 1.08 2.57 2.51 2.43 2.94 1.67 1.92 1.80 2.00 1.84 1.90 1.83 1.90 1.85 1.91 1.81 1.70 1.62
134
Behavioral repertoire14 (gets unit to meet goals) Behavioral repertoire15 (makes unit's roles clear) Behavioral repertoire16 (clarifies unit's priorities) Self-confidence in ldrship1 (confident ld in most sit.) Self-confidence in ldrship2 (lead others) Self-confidence in ldrship3 (lead group to perform) Self-confidence in ldrship4 (ability to influence) Self-confidence in ldrship 5 (to overcome obstacles) Self-confidence in ldrship6 (set direction) Self-confidence in ldrship7 (to gain commitment)
180 181 182 182 183 183 182 183 182 183
1 1 1 1 1 1 1 1 1 1
7 7 7 7 7 7 7 7 7 7
4.94 4.92 4.86 5.17 5.13 5.04 4.95 5.15 5.01 4.89
1.70 1.81 1.82 1.30 1.37 1.40 1.38 1.27 1.32 1.33
135
Appendix E Survey used for Student Sample
136
Perceptions About Leadership Before completing this survey, you must read the "Informed Consent" statement below ORP USE ONLY: The Pennsylvania State University Office for Research Protections
Informed Consent
Approval Date: 10/22/04 JKG
The Pennsylvania State University
Expiration Date: 10/18/05 JKG
Title of Project:
Perceptions about Leadership
Principal Investigator:
Nathan J. Hiller 429 Moore Building University Park, PA 16802 (814) 777-2840;
[email protected]
Advisor:
Dr. David V. Day 124 Moore Building University Park, PA 16802 (814) 865-3180;
[email protected]
1. Purpose of the Study: The purpose of this study is to better understand how different people understand leadership and how they perceive themselves in the domain of leadership. 2. Procedures to be followed: You will be asked to answer a series of survey questions about leadership, personality characteristics and previous experiences. 3. Discomforts and Risks: There are no risks in participating in this research beyond those experienced in everyday life. Some of the questions are personal and might cause discomfort. 4. Benefits: This research might help us better understand how to train and develop leadership in both individuals and groups. In addition, you might learn more about leadership and yourself by participating in this study. 5. Duration: It will take about 45-60 minutes to complete the survey. 6. Statement of Confidentiality: Your confidentiality will be maintained to the degree permitted by the technology used. Specifically, no guarantees can be made regarding the interception of data sent via the Internet by any third parties. Your course instructor will not see any information you provide other than that you participated in this research.
137 The data will be stored and secured on Nathan Hiller’s computer located in a locked office. The Office for Research Protections may review records related to this project. In the event of a publication or presentation resulting from the research, no personally identifiable information will be shared. 7. Right to Ask Questions: You can ask questions about this research. Contact Nathan Hiller at (814) 777-2840 with questions. If you have questions about your rights as a research participant, contact The Pennsylvania State University’s Office for Research Protections at (814) 865-1775. 8. Compensation: You will receive course credit for participating as specified in the syllabus provided by your instructor. In order to receive this credit, you will need to type your name and date in the spaces below. Alternative means for earning this credit are available as described in the syllabus. You are entitled to no other compensation. 9. Voluntary Participation: Your decision to be in this research is voluntary. You can stop at any time. You do not have to answer any questions you do not want to answer. You must be 18 years of age or older to take part in this research study. If you agree to take part in this research study and the information outlined above, please type your name and indicate the date below. 1. Please fill in the two spaces below. Name Date 2. Class Psy 002 (all classes)
Please print a copy of this consent form for your records. ** To avoid printing the entire document, please select all text above this box and copy and paste into a word processing package, then print that document.** After you have finished, complete the survey below
Question Set #1. Characteristics of Leaders
How characteristic of a typical leader is each of the following traits? Please answer the following items on a scale from 1 (not at all characteristic) to 5 (completely characteristic).
138 Not at all Usually not characteristic characteristic 1 2 Understanding Domineering Helpful Selfish Dynamic Conceited Funny Intelligent Manipulative Loud Knowledgeable Masculine Clever Educated Dedicated Sincere Motivated Hard-Working Energetic Male Cheerful Strong Pushy
Somewhat characteristic 3
Usually characteristic 4
Completely characteristic 5
139 Question #2.
About Yourself
Please answer the following questions about yourself on a scale from 1 (strongly disagree) to 7 (strongly agree)
Strongly Somewhat Disagree Disagree Disagree 1
2
Neither Agree nor Disagree
3
Somewhat Strongly Agree Agree Agree 5
6
7
4
I hate to disagree with others in my group. I often 'do my own thing'. I prefer to be direct and forthright when dealing with people. The well-being of other students I deal with is important to me. When another person does better than I do, I get tense. When I succeed, it is usually because of my abilities. Without competition, it is not possible to have a good society. Children should be taught to place duty before pleasure. I would do what would please my family, even if I detested that activity. I like sharing little things with my neighbors. I enjoy being unique and different from others. My happiness depends very much on the happiness of those around me. Winning is everything. It is important that I do my job better than others. It annoys me when other people perform better than I do.
1
2
3
4
5
6
7
140 I enjoy working in situations involving competition with others. Some people emphasize winning, I'm not one of them. If a student colleague gets a prize, I would feel proud.
1
2
3
4
5
6
7
If a relative were in financial difficulty, I would help within my means. It is important to maintain harmony within my group. I feel good when I cooperate with others. I would sacrifice an activity that I enjoy if my family did not approve of it. I usually sacrifice my self-interest for the benefit of my group. I am a unique individual. Competition is the law of nature. Children should feel honored if their parents receive a distinguished award. To me, pleasure is spending time with others. Strongly Disagree Somewhat Neither Somewhat Agree Strongly Agree Disagree Disagree Agree nor Agree Disagree
About Yourself Q3. How characteristic of you is each trait? Answer using the scale below, from 1 (not at all characteristic) to 5 (completely characteristic).
Not at all characteristic 1 Understanding
Usually not characteristic 2
Somewhat characteristic 3
Usually characteristic 4
Completely characteristic 5
141 Domineering Helpful Selfish Dynamic Conceited Funny Intelligent Manipulative Loud Knowledgeable Masculine Clever Educated Dedicated Sincere Motivated Hard-Working Energetic Male Cheerful Strong Pushy Not at all characteristic
About Yourself
Usually Not characteristic
Somewhat characteristic
Usually characteristic
Completely characteristic
142 Q4. Please answer the following items on a scale from 1 (strongly disagree) to 7 (strongly agree). Somewhat Neither Strongly agree Disagree Disagree Disagree nor disagree 1 2 3 4
Strongly Somewhat Agree Disagree Agree 5
6
7
I am confident I get the success I deserve in life. Sometimes I feel depressed. When I try, I generally succeed. Sometimes when I fail I feel worthless. I complete tasks successfully. Sometimes, I do not feel in control of my work. Overall, I am satisfied with myself. I am filled with doubts about my competence. I determine what will happen in my life. I do not feel in control of my success in my career. I am capable of coping with most of my problems. There are times when things look pretty bleak and hopeless to me.
Experiences as a leader. Think about your past leadership experiences. Q5. Please answer the following items on a scale of 1 (never) to 5 (all the time). Never Rarely Occasionally Most of Nearly all the time the time 1 2 3 4 5
143 When a leader is chosen, how often are you chosen as a leader? How often do you take on leadership responsibilities? How often do others look to you for leadership? How often are you involved in leadership, either formally or informally?
Q6. Please answer the following questions about the quality of your leadership experience, on a scale from 1 (strongly disagree) to 7 (strongly agree). Keep in mind that leadership does not have to involve a formal position of authority over others. Neither Strongly Somewhat Strongly Somewhat Agree Agree Disagree Disagree Disagree Agree nor Agree 2 Disagree 1 5 3 6 4 7 I received a significant amount of feedback from others about my leadership performance. I received information that helped me to see my strengths and weaknesses more clearly. I received useful advice from a person who had a position of higher authority than me. I could usually tell how effectively I was leading, even when nobody said anything. I was able to get information from the task itself that allowed me to understand how I was doing as a leader. I received useful information from my peers or a close friend about my leadership. I received feedback that helped me to lead more effectively.
Leadership Experiences Please answer the following questions in relation to yourself, using the scale below, using a scale from 1 (strongly disagree) to 7 (strong agree).
144 Q7. Over the course of my leadership experiences, in general... Strongly Disagree 1
Disagree
2
Somewhat Disagree
3
Neither Somewhat Agree Strongly Agree Agree nor Agree Disagree 5 6 7 4
Someone provided me with support as I developed as a leader. At least one person could be considered a mentor to me. I had the resources necessary to become a better leader. Even if I made mistakes, people around me were generally supportive and encouraging. I was allowed to further develop my leadership abilities and potential. I was encouraged to further develop my leadership abilities and potential.
Challenge and Intensity Q8. Please rate the challenge and intensity of your leadership experiences, on a scale from 1 (strongly disagree) to 7 (strongly agree). Neither Somewha Somewhat Strongly Agree Strongly Disagree Agree Disagree nor Disagree Agree Agree Disagree 2 6 1 3 5 7 4 I've had a leadership experience(s) that was (were) very challenging. When I look back on my leadership experiences, there was at least one experience that pushed me to the limits of my abilities. At least one of my leadership experiences involved reaching a difficult goal. I've had at least one leadership experience where I had to juggle competing demands. I felt like giving up at some point during a leadership experience. I've had a wide variety of different leadership experiences
145
Please take a few minutes to get up from the computer and stretch. You are over half way finished! :)
The Nature of Leadership Q9. The following items ask about your beliefs about the nature of leadership. Please respond using the scale below, from 1 (strongly disagree) to 7 (strongly agree). Strongly Disagree Somewhat Disagree
1
Telling other people what to do is a big part of leadership. Group members can work together in performing leadership tasks. Leaders order other people around. Leadership involves persuading other people. Effective leaders seek different perspectives. Individual people do not possess leadership -- it is a property of the group. Leaders should not let others openly question them. Leadership happens when people collaborate. The leader should be the most influential person in a group. Leadership is a process. Leadership and power are pretty much the same thing. All group members should participate in leadership. A leader never shows weakness. Leadership is about taking charge of a situation. People can be taught to be more effective leaders. Being a leader requires using both rewards and
Disagree 2
3
Strongly Neither Somewhat Agree Agree Agree nor Agree Disagree 4 5 6 7
146 punishment. Skills and abilities for leadership can be developed. Leadership is a property of the group, not the individual. Leadership is about maintaining control. Leadership involves a group collectively making decisions. Getting 'buy-in' from followers is an important leadership task. Effective leaders influence followers more than followers influence them. Leaders, as people, are fundamentally different from non-leaders. Leadership can be shared. Leadership is the responsibility of everybody in a group. When a group performs well, it is largely because of the leader. Leadership is about using influence. One's formal position determines whether they are a leader. Leadership is a two-way street between leaders and followers. Leaders can acquire skills to make them more effective. Together, group members create leadership. You're either a leader, or you're not. Leaders allow followers to influence them. You can't teach leadership. Leadership is about the group, rather than a single leader. Leadership is not possessed by any one individual. If you supervise others, you are a leader.
About Yourself
147 Q10. Please rate the extent to which the following statements describe you, using a scale from 1 (not at all descriptive) to 5 (extremely descriptive). Not at all Descriptive 1
Mostly Not Descriptive 2
Occasionally Descriptive 3
Mostly Totally Descriptive Descriptive 4 5
I am a leader. If I had to describe myself to others, I would include the word ''leader''. I prefer being seen by others as a leader. I see myself as a leader.
Q11. How certain are you about the ratings you gave for each statement above? Please rate from 1 (not at all certain) to 5 (extremely certain). Totally Mostly Somewhat Mostly Extremely Uncertain Uncertain Certain Certain Certain 1 2 3 4 5 I am a leader. I see myself as a leader. If I had to describe myself to others, I would include the word ''leader''. I prefer being seen by others as a leader.
Q12. Think about your overall self-concept. How important are each of the statements to your self-identity. Answer the following questions below, from 1 (not at all important) to 5 (extremely important). Not at all Mostly Somewhat Mostly Extremely Important Unimportant Important Important Important 1 2 3 4 5 I am a leader.
I see myself as a leader. If I had to describe myself to others, I would include the word ''leader''.
148
I prefer being seen by others as a leader.
About Yourself Q13. Please respond to the following statements about yourself. Use the scale below as an approximate guide.
0% 15% 35% 50% 65% 85% 100% confidence confidence confidence confidence confidence confidence confidence
I feel confident that I can be a leader in most situations. I am confident that I can lead others effectively. I am confident that I can lead a group to perform effectively. I am confident of my ability to influence a group I lead. I am confident that I can help group overcome obstacles. I am confident of my ability to set a direction for the group. I am confident of my ability to gain commitment from others.
About Yourself Q14. Please respond to the following statements, using a scale from 1 (strongly disagree) to 7 (strongly agree). Strongly Disagree 1
Most of the time, I prefer being a leader rather than a follower when working in a group. I would never agree to lead just because others voted for me.
Disagree 2
Somewhat Disagree 3
Neither Somewhat Strongly Agree Agree nor Agree Agree Disagree 6 5 7 4
149 I feel that I have a duty to lead others if I am asked. I would agree to lead others even if there are no special rewards or benefits with that role. I am the type of person who likes to be in charge of others. If I agree to lead a group, I would never expect any advantages or special benefits. I usually want to be the leader in the groups that I work in. It is not right to decline leadership roles. I have a tendency to take charge in most groups or teams that I work in.
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Leading others is really more of a dirty job rather than an honorable one. I am only interested to lead a group if there are clear advantages for me. I agree to lead whenever I am asked or nominated by the other members. I will never agree to lead if I cannot see any benefits from accepting that role. I am the type who would actively support a leader but prefers not to be appointed as leader. It is an honor and privilege to be asked to lead. I would only agree to be a group leader if I know I can benefit from that role. I would want to know ''what's in it for me'' if I am going to agree to lead a group. I am the type of person who is not interested to lead others. I have more of my own problems to worry about than to be concerned about the rest of the group. I was taught to believe in the value of leading others. I am seldom reluctant to be the leader of a group. It is appropriate for people to accept leadership roles or positions when they are asked.
150 I believe I can contribute more to a group if I am a follower rather than a leader. I have been taught that I should always volunteer to lead others if I can. I am definitely not a leader by nature. I never expect to get more privileges if I agree to lead a group. People should volunteer to lead rather than wait for others to ask or vote for them. Strongly Somewhat Disagree Disagree Disagree
Neither Agree Somewhat Strongly nor Agree Agree Agree Disagree
Display of Self Q15. Please answer the following items, using a scale from 1 (strongly disagree) to 7 (strongly agree).
Neither Strongly Disagree Somewhat Agree Somewhat Disagree Disagree nor Agree 1 3 5 Disagree 2 4
In social situations, I have the ability to alter my behavior if I feel that something else is called for. I am often able to read people's true emotions correctly through their eyes. I have the ability to control the way I come across to people, depending on the impression I wish to give them. In conversations, I am sensitive to even the slightest change in the facial expression of the person I'm conversing with. My powers of intuition are quite good when it comes to understanding others' emotions and motives. I can usually tell when others consider a joke to be in bad taste, even though they may laugh convincingly. When I feel that the image I am portraying isn't working, I can readily
Agree Strongly Agree 7 6
151 change it to something that does. I can usually tell when I've said something inappropriate by reading it in the listener's eyes. I have trouble changing my behavior to meet the requirements of any situation I find myself in. I have found that I can adjust my behavior to meet the requirements of any situation I find myself in. If someone is lying to me, I usually know it at once from that person's manner of expression. Even when it might be to my advantage, I have difficulty putting up a good front. Once I know what the situation calls for, it's easy for me to regulate my actions accordingly. Strongly Disagree Somewhat Disagree Disagree
Neither Agree Somewhat Agree Strongly nor Agree Agree Disagree
When you are satisfied that you have answered all items carefully, please click the "Submit Survey" button below. Submit Survey
Please Note: When you click "Submit Survey", it will take up to several minutes for the "Thank You" confirmation page to appear. Do not close your browser until you see this "Thank You" page.
This questionnaire was created using Perseus SurveySolutions.
152
Appendix F Factor Analysis Results of Leadership Experience Measure
153 Leadership Experience Measure: Results of Principal Axis Factoring with Varimax Rotation Factor Items 1. Quantity of Experience1: How often chosen leader 2. Quantity of Experience2: take on leadership
1 .141 .145
2 .794 .764
3 .111 .082
4 .136 .182
3. Quantity of Experience3: often others look to you 4. Quantity of Experience4: often involved in leadership
.097 .145
.623 .812
.110 .147
.148 .192
5. Feedback about Leadership1: significant amount of feedback 6. Feedback about Leadership2: access to info about strengths and weaknesses 7. Feedback about Leadership3: useful advice from higher up
.255
.279
.626
.094
.288
.143
.727
.117
.300
-.003
.502
.145
8. Feedback about Leadership4: I could tell how effective I was
.187
.248
.189
-.012
9. Feedback about Leadership5: received task feedback
.220
.158
.260
.107
10. Feedback about Leadership6: received feedback from peers 11. Feedback about Leadership7: received feedback to help me lead more effectively 12. Developmental Support1: support as I developed
.243
.140
.751
.144
.294
.129
.762
.161
.632
.111
.375
.195
13. Developmental Support2: at least one mentor
.626
.004
.255
.131
14. Developmental Support 3: access to resources necessary to be better
.721
.167
.283
.138
15. Developmental Support4: support when I made mistake 16. Developmental Support5: allowed to develop abilities and pot 17. Developmental Support6: encouraged to develop 18, Challenging Experiences1: had challenging experiences 19. Challenging Experiences2: had one experience that pushed me to limit 20. Challenging Experiences3: had one experience reaching a difficult goal 21. Challenging Experience4: juggle competing demands
.661
.123
.159
-.013
.759
.260
.226
.101
.753 .220
.230 .309
.159 .172
.137 .607
.120
.235
.186
.740
.265
.196
.149
.637
.221
.306
.164
.561
154 22. Challenging Experiences5: felt like giving up at some point 23. Challenging Experiences6: wide variety of ldrship experiences
-.056
.013
-.008
.360
.194
.523
.250
.426
Note. Bold items indicate clear loading on intended factor. N = 441.
155
Appendix G Pre-Survey Letter Sent to PSHMC Participants
156
June 8, 2005
Dear PSHMC employee, I am writing to ask you to participate in a research study that may help to better understand the phenomenon of leadership, both generally, and here at the Penn State Milton S. Hershey Medical Center. This research is being conducted by myself (Dr. David Day) and Nathan Hiller at the University Park campus, and Dr. Chip Souba of PSHMC. This letter has been sent by the Human Resources Department at PSHMC to clinical faculty and nurses, division chiefs/chairs, and nurse managers/clinical head nurses. The Human Resources Department is overseeing distribution of surveys in order to maintain the confidentiality of participant addresses and contact information. The only information investigators have been provided with is a pairing of supervisor/supervisee random codes. A few days from now you will receive a brief survey in the mail, along with a description and instructions. You can participate in this research by completing the survey and mailing it to the confidential collection site at the University Park campus. This information will be used to deepen our understanding of how to make leadership work effectively at PSHMC. In all, approximately 1000 individuals have been selected to participate in this study. In units where a ‘supervisor’ has more than 10 direct reports, 10 individuals will be randomly chosen. In units where a supervisor has less than 10 direct reports, all individuals will be selected for inclusion in this study. Your responses will remain confidential; no one at Hershey (including Dr. Souba) will see any individual results. General trends and findings will be made available to PSHMC in several months. Thank you for your time and consideration. It’s only with the help of everyone that this research can be successful. Sincerely,
David Day, Ph.D. Professor of Organizational Psychology Penn State University, Main Campus (814) 865-3180
Dr. Wiley “Chip” Souba Chair, Department of Surgery Penn State Hershey Medical Center (717) 531-8939
157
Appendix H Survey used for PSHMC Supervisor Sample
158 Penn State Hershey Medical Center
Leadership Survey
Code: _____________
4
5
6
7
3
4
5
6
7
3. Effective leaders seek different perspectives.
1
2
3
4
5
6
7
4. Leadership and power are pretty much the same thing.
1
2
3
4
5
6
7
5. One’s formal position determines whether he or she is a leader.
1
2
3
4
5
6
7
6. Leadership involves persuading other people.
1
2
3
4
5
6
7
7. All group members should participate in leadership.
1
2
3
4
5
6
7
8. Being a leader requires using both rewards and punishment.
1
2
3
4
5
6
7
9. Leaders can acquire skills to make them more effective.
1
2
3
4
5
6
7
10. Leadership is about maintaining control.
1
2
3
4
5
6
7
11. People are either leaders, or they're not.
1
2
3
4
5
6
7
12. People can't be taught leadership.
1
2
3
4
5
6
7
13. Leaders, as people, are fundamentally different from non-leaders.
1
2
3
4
5
6
7
14. Individual people do not possess leadership -- it is a property of the group.
1
2
3
4
5
6
7
15. Leadership is about taking charge of a situation.
1
2
3
4
5
6
7
16. Leadership is about using influence.
1
2
3
4
5
6
7
17. Leadership is a process.
1
2
3
4
5
6
7
18. Leadership can be shared.
1
2
3
4
5
6
7
19. People can be taught to be more effective leaders.
1
2
3
4
5
6
7
20. Leadership is the responsibility of everybody in a group.
1
2
3
4
5
6
7
21. Effective leaders influence followers more than followers influence them.
1
2
3
4
5
6
7
22. Leaders order other people around.
1
2
3
4
5
6
7
23. The leader should be the most influential person in a group.
1
2
3
4
5
6
7
24. Leadership involves a group collectively making decisions.
1
2
3
4
5
6
7
25. Leaders should minimize the amount that others question them.
1
2
3
4
5
6
7
26. Leadership is about the group, rather than a single leader.
1
2
3
4
5
6
7
27. Leadership happens when people collaborate.
1
2
3
4
5
6
7
28. When a group performs well, it is mainly because of the leader.
1
2
3
4
5
6
7
29. Leadership is not possessed by any one individual.
1
2
3
4
5
6
7
30. If you supervise others, you are a leader.
1
2
3
4
5
6
7
31. Leadership is a property of the group, not the individual.
1
2
3
4
5
6
7
32. Skills and abilities for leadership can be developed.
1
2
3
4
5
6
7
33. Leadership is a two-way street between leaders and followers.
1
2
3
4
5
6
7
34. A leader never shows weakness.
1
2
3
4
5
6
7
35. Group members can work together in performing leadership tasks.
1
2
3
4
5
6
7
36. Together, group members create leadership.
1
2
3
4
5
6
7
37. Leaders allow followers to influence them.
1
2
3
4
5
6
7
Strongly Agree
3
2
Agree
Somewhat Agree
2
1
Somewhat Disagree
1
2. Getting ‘buy-in’ from followers is an important leadership task.
Disagree
1. Telling other people what to do is a big part of leadership.
Your Personal Views about Leadership in General
Strongly Disagree
Neither Agree nor Disagree
Please complete the following sets of items by placing an 'X' in the appropriate circle.
Please answer the following questions about your views of leadership in general.
159 Perceptions about Yourself Please read the following statements.
1. 2. 3. 4.
"I am a leader" "I see myself as a leader" "If I had to describe myself to others, I would include the word 'leader' " "I prefer being seen by others as a leader"
A. How descriptive is each statement of you? Moderately Descriptive.
Not at all Descriptive
Statement 1 Statement 2 Statement 3 Statement 4
Extremely Descriptive
. .
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
. .
B. How certain are you about the ratings you just gave above? Not at all Certain
Statement 1 Statement 2 Statement 3 Statement 4
Moderately Certain .
. .
Extremely Certain
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
. .
C. How important to you is this view of yourself? Not at all Important
. .
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Disagree
Somewhat Disagree
Agree
Strongly Agree
Extremely Important
Strongly Disagree
Statement 1 Statement 2 Statement 3 Statement 4
Moderately Important .
1. If given an opportunity, I would participate in a leadership development program. 2. I have a strong desire to develop as a leader. 3. I continually look for ways to develop my leadership potential.
Somewhat Agree
Please answer the following statements about yourself.
Neither Agree nor Disagree
Development of Leadership
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Continue on next page.
. .
4
5
6
7
3
4
5
6
7
3. Subordinates can question their superiors on important issues.
1
2
3
4
5
6
7
4. People feel involved with the process of leadership.
1
2
3
4
5
6
7
5. The mission is aligned with the goals of the institution.
1
2
3
4
5
6
7
6. People provide honest feedback to others.
1
2
3
4
5
6
7
7. People are not held accountable for their performance.
1
2
3
4
5
6
7
8. Constructive conflict is embraced and encouraged.
1
2
3
4
5
6
7
9. Communication is driven by hidden agendas. 10. Information is widely shared across departments. 11. Everyone can be a leader. 12. Good performance is rewarded. 13. People do not collaborate effectively in teams. 14. Constructive conflict is not tolerated. 15. Decision-making is transparent to those at lower levels. 16. Leaders are held accountable for their performance. 17. People do not communicate openly.
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Strongly Agree
3
2
Agree
Somewhat Agree
2
1
Somewhat Disagree
1
2. People are kept in the dark about information and decisions.
Climate for Leadership
Disagree
1. Leadership is widely shared.
Strongly Disagree
Neither Agree nor Disagree
160
5
6
7
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Strongly Agree
4
3
Agree
Somewhat Agree
3
2
Somewhat Disagree
2
1
Disagree
1
Strongly Disagree
Neither Agree nor Disagree
At Penn State Hershey Medical Center (PSHMC):
At Penn State Hershey Medical Center (PSHMC): 18. Performance expectations are clear. 19. Decisions are made in secret. 20. People feel open to disagree with others on key issues. 21. People would give similar answers to the question "What are we trying to accomplish here?" 22. Everyone is encouraged to participate in leadership. 23. Departments have their own agendas, which take priority over the institutional mission. 24. Values are aligned to meet the needs of patients or other key stakeholders. 25. People work together in solving problems. 26. Poor performers are held accountable. 27. Competition is the norm between departments. 28. Information is tightly held by those in power. 29. People collaborate productively. 30. Leadership is the responsibility of only those at the top. 31. Teamwork is practiced widely. 32. People understand that conflict and debate produce better ideas. 33. Open, frank communication is the norm. 34. People share similar values. 35. Open communication is valued. Please provide us with the following personal information: 1. What is your age group? (please circle) 20-30 yrs,
31-40 yrs, 41-50 yrs, 51-60 yrs,
2. How long have you been in your current position at PSHMC?
______ yrs.
over 60 yrs
161
Please answer the following questions about each employee who's name is listed on the following pages. If you supervise more than 10 individuals, 10 individuals were randomly selected. A pre-coded sheet is provided for each employee with a strip of paper attached for temporary identification purposes. Please peal off and throw away the attached strip(s) of paper once you are finished.
Relationship with Individual Faculty/ Nurse Code ____________
Please answer the following questions about the employee who's name is attached. Remove paper when complete. 1. Does he/she know how satisfied you are with the work he/she does? Rarely
Occasionally
1
2
Sometimes
Fairly often
3
Very often
4
5
2. How well do you understand the job problems and needs of this employee? Not at all
1
A little
2
A fair amount
Quite a bit
3
A great deal
4
5
3. How well do you recognize the potential of this employee? Not at all
1
A little
2
Moderately
3
Mostly
4
Fully
5
4. Regardless of how much formal authority you have, what are the chances that you would use your power to help solve work-related problems for this employee? None
1
Small
Moderate
2
3
High
Very high
4
5
5. Again, regardless of the amount of formal authority you have, what are the chances that you would 'bail' an employee out at your None 1 Small 2 Moderate 3 High 4 Very high 5 expense?
6. This employee has enough confidence in me that he/she would justify my decision to others if I were not present to do so. Strongly Disagree
Disagree
1
2
Neutral
3
Agree
4
Strongly Agree
5
7. How would this employee characterize their working relationship with you? Extremely ineffective
1
Worse than average
2
Average
3
Above average
Extremely Effective
4
5
Leadership Potential of This Faculty/Nurse 0% Confidence
1. How much confidence do you have in his/her overall leadership ability? 2. How much confidence do you have in his/her ability to lead others? 3. How much confidence do you have in his/her potential to learn to be an effective leader? 4. How much confidence do you have that this person would be an effective department chair/division chief/nurse manager/clinical head?
. .
Moderate Confidence
100% Confidence
0
1
2
3
4
5
6
7
8
9
10
0
1
2
3
4
5
6
7
8
9
10
0
1
2
3
4
5
6
7
8
9
10
0
1
2
3
4
5
6
7
8
9
10
162
Appendix I Survey Used for PSHMC Subordinate Sample
163 Penn State Hershey Medical Center
Leadership Survey
Code: _____________
4
5
6
7
2. Getting ‘buy-in’ from followers is an important leadership task.
1
2
3
4
5
6
7
3. Effective leaders seek different perspectives.
1
2
3
4
5
6
7
4. Leadership and power are pretty much the same thing.
1
2
3
4
5
6
7
5. One’s formal position determines whether he or she is a leader.
1
2
3
4
5
6
7
6. Leadership involves persuading other people.
1
2
3
4
5
6
7
7. All group members should participate in leadership.
1
2
3
4
5
6
7
8. Being a leader requires using both rewards and punishment.
1
2
3
4
5
6
7
9. Leaders can acquire skills to make them more effective.
1
2
3
4
5
6
7
10. Leadership is about maintaining control.
1
2
3
4
5
6
7
11. People are either leaders, or they're not.
1
2
3
4
5
6
7
12. People can't be taught leadership.
1
2
3
4
5
6
7
13. Leaders, as people, are fundamentally different from non-leaders.
1
2
3
4
5
6
7
14. Individual people do not possess leadership -- it is a property of the group.
1
2
3
4
5
6
7
15. Leadership is about taking charge of a situation.
1
2
3
4
5
6
7
16. Leadership is about using influence.
1
2
3
4
5
6
7
17. Leadership is a process.
1
2
3
4
5
6
7
18. Leadership can be shared.
1
2
3
4
5
6
7
19. People can be taught to be more effective leaders.
1
2
3
4
5
6
7
20. Leadership is the responsibility of everybody in a group.
1
2
3
4
5
6
7
21. Effective leaders influence followers more than followers influence them.
1
2
3
4
5
6
7
22. Leaders order other people around.
1
2
3
4
5
6
7
23. The leader should be the most influential person in a group.
1
2
3
4
5
6
7
24. Leadership involves a group collectively making decisions.
1
2
3
4
5
6
7
25. Leaders should minimize the amount that others question them.
1
2
3
4
5
6
7
26. Leadership is about the group, rather than a single leader.
1
2
3
4
5
6
7
27. Leadership happens when people collaborate.
1
2
3
4
5
6
7
28. When a group performs well, it is mainly because of the leader.
1
2
3
4
5
6
7
29. Leadership is not possessed by any one individual.
1
2
3
4
5
6
7
30. If you supervise others, you are a leader.
1
2
3
4
5
6
7
31. Leadership is a property of the group, not the individual.
1
2
3
4
5
6
7
32. Skills and abilities for leadership can be developed.
1
2
3
4
5
6
7
33. Leadership is a two-way street between leaders and followers.
1
2
3
4
5
6
7
34. A leader never shows weakness.
1
2
3
4
5
6
7
35. Group members can work together in performing leadership tasks.
1
2
3
4
5
6
7
36. Together, group members create leadership.
1
2
3
4
5
6
7
37. Leaders allow followers to influence them.
1
2
3
4
5
6
7
Strongly Agree
3
Agree
Somewhat Agree
2
Somewhat Disagree
1
Disagree
1. Telling other people what to do is a big part of leadership.
Your Personal Views about Leadership in General
Strongly Disagree
Neither Agree nor Disagree
Please complete the following sets of items by placing an 'X' in the appropriate circle.
Please answer the following questions about your views of leadership in general.
164 Relationship with your Supervisor Please answer the following questions about your relationship with your supervisor named on the strip of paper. For confidentiality purposes, please tear off and throw away the strip of paper attached once you are finished. 1. Do you know how satisfied your leader is with what you do? Rarely 1 Occasionally 2 Sometimes 3 Fairly often 4 Very often
5
2. How well does your leader understand your job problems and needs? Not at all 1 A little 2 A fair amount 3 Quite a bit 4 A great deal 3. How well does your leader recognize your potential? Not at all 1 A little 2 Moderately 3 Mostly
4
Fully
5
5
4. Regardless of how much formal authority he/she has build into his/her position, what are the chances that your leader would use his/her power to help you solve problems in your work? None 1 Small 2 Moderate 3 High 4 Very high 5 5. Again, regardless of the amount of formal authority your leader has, what are the chances that he/she would ‘bail you out,’ at his/her expense? None 1 Small 2 Moderate 3 High 4 Very high 5 6. I have enough confidence in my leader that I would defend and justify his/her decision if he/she were not present to do so. Strongly Disagree 1 Disagree 2 Neutral 3 Agree 4 Strongly Agree 5 7. How would you characterize your working relationship with your leader? Extremely ineffective 1 Worse than average 2 Average Above average 3
Extremely Effective
4
5
Your Views of your Supervisor Please answer the following items about the supervisor listed on the strip of paper above.
Never
My Supervisor: 1. Treats each individual in a sensitive, caring way.
Once in a While
Sometimes
Fairly Often
Often
Most of the Time Always
1
2
3
4
5
6
7
2. Encourages participative decision-making in the group.
1
2
3
4
5
6
7
3. Shows empathy and concern in dealing with subordinates.
1
2
3
4
5
6
7
4. Works participatively to resolve differences among group members.
1
2
3
4
5
6
7
5. Experiments with new concepts and ideas.
1
2
3
4
5
6
7
6. Comes up with inventive ideas.
1
2
3
4
5
6
7
7. Influences decisions made at higher levels.
1
2
3
4
5
6
7
8. Exerts upward influence in the organization.
1
2
3
4
5
6
7
Sometimes
Fairly Often
Often
Never
Once in a While
Most of the Time Always
My Supervisor: 9. Anticipates workflow problems in order to avoid crises.
1
2
3
4
5
6
7
10. Brings a sense of order into the unit.
1
2
3
4
5
6
7
11. Maintains tight logistical control.
1
2
3
4
5
6
7
12. Compares records, reports, and so on, to detect discrepancies.
1
2
3
4
5
6
7
13. Sees that the unit delivers on stated goals.
1
2
3
4
5
6
7
14. Gets the unit to meet expected goals.
1
2
3
4
5
6
7
15. Makes the unit’s role very clear.
1
2
3
4
5
6
7
16. Clarifies the unit’s priorities and directions.
1
2
3
4
5
6
7
165 Perceptions about Yourself Please read the following statements.
1. 2. 3. 4.
"I am a leader" "I see myself as a leader" "If I had to describe myself to others, I would include the word 'leader' " "I prefer being seen by others as a leader"
A. How descriptive is each statement of you? Moderately Descriptive.
Not at all Descriptive
Statement 1 Statement 2 Statement 3 Statement 4
Extremely . Descriptive.
. .
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
B. How certain are you about the ratings you just gave above? Moderately Certain .
Not at all Certain
Statement 1 Statement 2 Statement 3 Statement 4
. .
Extremely Certain .
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
C. How important to you is this view of yourself? Extremely . Important .
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Agree
Strongly Agree
. .
Somewhat Disagree
Statement 1 Statement 2 Statement 3 Statement 4
Moderately Important .
Disagree
Not at all Important
Somewhat Agree
1. If given an opportunity, I would participate in a leadership development program. 2. I have a strong desire to develop as a leader. 3. I continually look for ways to develop my leadership potential.
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Strongly Disagree
Please answer the following questions about yourself.
Neither Agree nor Disagree
Development of Leadership
Leadership Beliefs about Self Please answer the following questions about yourself. 1. 2. 3. 4. 5. 6. 7.
I feel confident that I can be a leader in most situations. I am confident that I can lead others effectively. I am confident that I can lead a group to perform effectively. I am confident of my ability to influence a group I lead. I am confident that I can help a group overcome obstacles. I am confident of my ability to set direction for a group. I am confident of my ability to gain commitment from others.
0% Confidence
Moderate . Confidence.
100% Confidence
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
166
4
5
6
7
3
4
5
6
7
3. Subordinates can question their superiors on important issues.
1
2
3
4
5
6
7
4. People feel involved with the process of leadership.
1
2
3
4
5
6
7
5. The mission is aligned with the goals of the institution.
1
2
3
4
5
6
7
6. People provide honest feedback to others.
1
2
3
4
5
6
7
7. People are not held accountable for their performance.
1
2
3
4
5
6
7
8. Constructive conflict is embraced and encouraged.
1
2
3
4
5
6
7
9. Communication is driven by hidden agendas. 10. Information is widely shared across departments. 11. Everyone can be a leader. 12. Good performance is rewarded. 13. People do not collaborate effectively in teams. 14. Constructive conflict is not tolerated. 15. Decision-making is transparent to those at lower levels. 16. Leaders are held accountable for their performance. 17. People do not communicate openly.
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Strongly Agree
3
2
Agree
Somewhat Agree
2
1
Somewhat Disagree
1
2. People are kept in the dark about information and decisions.
Disagree
1. Leadership is widely shared.
Strongly Disagree
Neither Agree nor Disagree
Climate for Leadership
5
6
7
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
7
Strongly Agree
4
3
Agree
Somewhat Agree
3
2
Somewhat Disagree
2
1
Disagree
1
Strongly Disagree
Neither Agree nor Disagree
At Penn State Hershey Medical Center (PSHMC):
At Penn State Hershey Medical Center (PSHMC): 18. Performance expectations are clear. 19. Decisions are made in secret. 20. People feel open to disagree with others on key issues. 21. People would give similar answers to the question "What are we trying to accomplish here?" 22. Everyone is encouraged to participate in leadership. 23. Departments have their own agendas, which take priority over the institutional mission. 24. Values are aligned to meet the needs of patients or other key stakeholders. 25. People work together in solving problems. 26. Poor performers are held accountable. 27. Competition is the norm between departments. 28. Information is tightly held by those in power. 29. People collaborate productively. 30. Leadership is the responsibility of only those at the top. 31. Teamwork is practiced widely. 32. People understand that conflict and debate produce better ideas. 33. Open, frank communication is the norm. 34. People share similar values. 35. Open communication is valued. Please provide us with the following personal information: 1. What is your age group? (please circle) 20-30 yrs,
31-40 yrs, 41-50 yrs, 51-60 yrs,
2. How long have you been in your current position at PSHMC?
______ yrs.
over 60 yrs
167
Appendix J Summary Explanation of Research (Informed Consent) for PSHMC Participants
168 IRB Protocol # 20331 EP
Summary Explanation of Research Penn State College of Medicine Penn State Milton S. Hershey Medical Center Title of Project: Leadership Self-Identity and Personal Definitions of Leadership at PSHMC among Clinical Faculty, Division Chiefs/Chairs, Nurses, and Nurse Managers/Clinical Head Nurses.
Principal Investigator:
Other Investigators:
Dr. Wiley Souba Chair, Department of Surgery H051 Main Building, PSHMC Hershey, PA 17033-0850
David Day, Ph.D. & Nathan Hiller Dept. of Psychology, Penn State 417 Moore Building University Park, PA 16802
The purpose of this study is to better understand how clinical faculty, division chiefs, nurses, and clinical head nurses, understand leadership and how they perceive themselves with regard to leadership. You will be asked to answer a series of survey questions about your personal leadership views as well as the leadership atmosphere at PSHMC. This research might help us better understand how to train and develop leadership in general, and at PSHMC. Results will not be used to change official policies. In addition, you might gain a greater awareness of how you think about leadership in general and yourself as a leader. Strict confidentiality will be maintained during this process. Only Dr. David Day and Nathan Hiller (University Park campus) will see individual responses. A coding scheme has been devised so that your name and personal information are not recorded by the researchers, nor permanently attached to your individual responses. Dr. Souba will have access only to aggregated (general) data, free from information that could identify an individual or a department/unit. Data will be stored in a locked room at the University Park Campus (Room 125, Moore Building). In the event of a publication or presentation resulting from the research, no personally identifiable information will be shared. While the possibility for an unintended breach in confidentiality still exists, the procedures in place limit that risk as much as possible. In order for us to understand supervisor/employee relationships better, we will ask clinical faculty and nurses several questions about their relationship with, and leadership behaviors of, their supervisor. In addition, we will ask division chiefs (or chairs) and nurse managers (or clinical head nurses) about relationships with, and leadership potential of, each individual that they supervise. In units where a ‘supervisor’ has more than 10 supervisees, 10 individuals will be randomly chosen. In units where a supervisor has less than 10 supervisees, all individuals will be selected for inclusion in this study. If either individual…
169 …in this ‘pair’ does not return a survey, the remaining data for that supervisor/supervisee pair will be removed. In order to link responses between supervisors and supervisees, surveys are pre-coded using a numerical code. In approximately seventeen days, the numerical coding sheet connecting the assigned code to individual names will be destroyed to protect the confidentiality of your responses. In addition to the risk of an unintended breach in confidentiality, other potential risks include: psychological discomfort in providing opinions, discomfort from knowing that your supervisor/supervisee is being asked questions about you, and the possibility that a supervisor might take retaliatory action if a faculty member or nurse made their opinions known. You do not have to participate in this research. Taking part in the research study is voluntary. Your decision to participate in or to decline the research will not affect your job status. You will not receive compensation or incur any expense for participation in this research. It will take about 15-20 minutes to complete the questions. Once you have completed the survey, please return it to the University Park campus using the enclosed postage-paid envelope. You have the right to ask any questions you may have about this research. If you have any questions, contact Dr. Wiley Souba (717-531-8939). If you have questions about your rights as a research participant or concerns regarding your privacy, you may contact the research protection advocate in the HMC Human Subjects Protection Office at 717-531-5687. The institution and investigators are not receiving any funding to support this research study. Your completion of the questionnaire implies your voluntary consent to participate in the research. This research is not connected in any way to your annual performance evaluations or the employee satisfaction survey.
170
Appendix K Follow-up Letter Sent to PSHMC Participants
171 Penn State Dept. of Psychology Letterhead June 24, 2004 Dear Penn State Hershey Medical Center Employee, About one week ago, we sent a questionnaire to you that asked about your leadership views and about the climate for leadership at PSHMC. If you have completed this survey, we thank you for your time and effort. If you have not completed and returned the survey, we would appreciate if you would send in your response by Friday July 1, 2005 All responses are confidential and are being collected by leadership researchers David V. Day and Nathan J. Hiller from the University Park Campus (Department of Psychology). Dr. Wiley “Chip” Souba at PSHMC, who is also involved in the project, will have access to compiled general information only. Access to individual survey responses will be limited to Dr. Day and Nathan Hiller and will be stored in a manner that will not allow identification of specific individuals or departments. General trends and results will be presented to PSHMC in several months. The results from this survey will advance our scientific understanding of leadership in general, and help to understand how to make leadership work more effectively at PSHMC. If you have any questions, or if you have misplaced your survey, please call Nathan Hiller at (814) 777-2840, or email
[email protected]. Sincerely,
David Day, Ph.D. Professor of Organizational Psychology Penn State University, Main Campus (814) 865-3180
Wiley “Chip” Souba, M.D., Sc.D. Chair, Department of Surgery Penn State Hershey Medical Center (717) 531-8939
Vita Nathan J. Hiller Education Ph.D. in Psychology, Industrial/Organizational (December, 2005) The Pennsylvania State University (Chair: David Day) Dissertation Title : An Examination of Leadership Beliefs and Leadership Self-Identity: Constructs, Correlates, and Outcomes M.S. in Psychology (December 2001) The Pennsylvania State University (Chair: David Day) Thesis Title: ‘Understanding and Measuring Shared Leadership in Work Teams.’ B.A. (Honours), Industrial/Organizational (May,1998) University of Calgary Publications Hiller, N. J., Day, D. V., & Vance, R. J. (in press). Collective enactment of leadership roles and team effectiveness: A field study. The Leadership Quarterly. Hiller, N. J., & Hambrick, D. C. (2005). Conceptualizing executive hubris: The role of (hyper-) core self-evaluations in strategic decision-making. Strategic Management Journal, 26, 297-319. Hiller, N. J., & Day, D. V. (2003). Leader-Member Exchange and teamwork: The challenges and opportunities of diversity. In G. Graen (Ed.), Dealing with Diversity: LMX, the Series, Vol. 1 (pp. 29-52). Greenwich, CT: Infoage. Day, D.V., Schleicher, D. J., Unckless, A. L., & Hiller, N. J. (2002). Self-monitoring personality at work: A meta-analysis. Journal of Applied Psychology, 87, 390-401. Hiller, N. J., & Kline, D. W. (2001). Diminished spatial summation contributes to the age deficit in the discrimination of low-contrast vernier oscillation. Optometry & Vision Science, 78, 616-622. Awards, Achievements, and Experience Crawford Dissertation Fellowship for 2003-2004 academic year. Research Assistant - Center for Applied Behavioral Science (1999-2003). Kenneth E. Clark Leadership Research Award, sponsored by the Center for Creative Leadership (2001). Applied Research/Consulting Experience: two major airlines, a manufacturing firm, Pennsylvania Department of Transportation, Center for Creative Leadership, United States Marine Corps.
