Items 1 - 6 - ings, designing a good rtsumt, filling out a job application, and visiting an ..... capture the computer-based job-search activity that has become more.
JotLrnal of Applied Psychology 2000, Vol. 85, No. 4, 491-503
Copyright 2000 by the American Psychological Association, Inc. 0021-9010/00/$5.00 DOI: 10.1037//0021-9010.85.4.491
Predictors and Outcomes of Networking Intensity Among Unemployed Job Seekers Connie R. Wanberg
Ruth Kanfer
University of Minnesota, Twin Cities Campus
Georgia Institute of Technology
Joseph T. Banas Washington University
This study examined predictors and outcomes of networking intensity (i.e., individual actions directed toward contacting friends, acquaintances, and referrals to get information, leads, or advice on getting a job) during the job searches of a sample of unemployed individuals. The study used a Big Five framework, in which extraversion and conscientiousness were associated with both higher levels of networking intensity and higher use of other traditional job-search methods. Networking comfort (a procedure-specific constellation of evaluative beliefs depicting attitudes toward using networking as a job-search method) was positively related to networking intensity above and beyond the effects of personality. Networking intensity did not provide incremental prediction of unemployment insurance exhaustion, reemployment or reemployment speed, or job satisfaction when intensity of use of other job-search methods was considered.
The economy in the United States, as indexed by unemployment rates, has noticeably improved in the past 5 years. For example, the seasonally adjusted unemployment rate in the United States for the civilian labor force was 7.3% in January of 1993, in contrast to 4.7% in January of 1998 (Bureau of Labor Statistics, 1998). Yet, despite this decrease in unemployment rate, the need for individuals to have job-search skills remains critical. Large layoffs are still common, and employees are also voluntarily transitioning between jobs at an increasing pace (Price & Vinokur, 1995). Welfare-to-work initiatives further amplify the need for a better understanding of job-search behavior. Finally, labor market shortages in several cities increase the need for an understanding of how job searches might be conducted to facilitate faster worker-job matches to remove individuals from unemployment ranks and into the job market more quickly.
Existing literature on the job search-reemployment relationship demonstrates the importance of job-search intensity (i.e., the frequency and scope of engagement in job-search behaviors) in reemployment success, as well as some of the key factors related to search intensity among employed and unemployed individuals. Among unemployed individuals, higher levels of economic hardship, employment commitment, conscientiousness, job-search self-efficacy, and support from significant others are associated with higher levels of job-search intensity (cf. R. Kanfer & Hulin, 1985; Rowley & Feather, 1987; Wanberg, Watt, & Rumsey, 1996). Among employed individuals, lower job satisfaction and commitment, higher ease of movement, and expected positive utility of changing jobs are associated with higher levels of job-search intensity (Blau, 1993). To date, job search-reemployment research has focused largely on job-search behaviors at the aggregate level, assessing jobsearch intensity or job-search frequency as the sum of the use of a variety of job-search methods in a set period of time. Few studies have examined the use of specific job-search methods. In this study, we focused on one particular job-search method, namely, networking. Specifically, our study examined a theoretically derived set of predictors (personality and networking comfort) and outcomes (exhaustion of unemployment insurance, reemployment, reemployment speed, and reemployment satisfaction) of networking use during the job search of a sample of unemployed individuals. Although employed individuals also use networking in their job searches, unemployed job seekers were the focus of this initial study of networking, because the job search tends to be a central priority among these individuals. Our focus on predictors and outcomes of networking as a job-search method arises from the emphasis on the importance of networking in current job-search books (e.g., Krannich & Krannich, 1996, Lowstuter & Robertson, 1995) and books on increas-
Connie R. Wanberg, Industrial Relations Center, University of Minnesota, Twin Cities Campus; Ruth Kanfer, Department of Psychology, Georgia Institute of Technology; Joseph T. Banas, John M. Olin School of Business, Washington University. This research was supported in part by a McKnight Business and Economics Research Grant and by National Science Foundation Grant NSF/SBR9223357. A special thank you goes to Jim Hegman from the Minnesota Department of Economic Security for his helpfulness in coordinating this statewide project. An article examining newcomer socialization behaviors among a subsample of reemployed individuals from this study was published in the June 2000 issue of the Journal of Applied Psychology, authored by Connie R. Wanberg and John D. KammeyerMueller. Correspondence concerning this article should be addressed to Connie R. Wanberg, Industrial Relations Center, University of Minnesota, Twin Cities CampuS, 3-255 Carlson School of Management, 321 19th Avenue South, Minneapolis, Minnesota 55455. Electronic mall may be sent to cwanberg @csom.umn.edu. 491
492
WANBERG, KANFER, AND BANAS
ing one's general career success (e.g., Baker, 1994). Supportive of this emphasis, research on job search and reemployment suggests that a large proportion of jobs are found through contacting friends, family, or other acquaintances or contacts (Granovetter, 1995; Schwab, Rynes, & Aldag, 1987). Reid (1972), for example, reported that approximately one third of 876 engineering and metal-trade workers found jobs through friends and family. Rosenfeld (1975) reported that although 11-35% of jobs in all occupational groups were filled through newspaper ads and employment agencies, 50-75% were filled by direct application or information or assistance provided by friends and relatives. Similarly, Granovetter (1995) estimated that 60-90% of blue-collar placements in the United States result from contacts with family or friends or direct application to employers. Findings in the jobchoice literature further indicate that informal methods of job search such as networking may also play a key role in employment among graduating college students (see, e.g., Allen & Keaveny, 1980; Stevens, Timaner, & Turban, 1997; Swaroff, Barclay, & Bass, 1985) and in job markets where positions typically are not advertised in public media (Meyer & Shadle, 1994). From an organizational perspective, several studies have indicated that employees hired through networking and employee referrals tend to have more positive work-related attitudes (Granovetter, 1995; Latham & Leddy, 1987), to stay longer with the organization (Taylor, 1994), and often to perform better (Breangh & Mann, 1984), although these results are not unequivocal (cf. Blau, 1990). A recent study by Griffeth, Horn, Fink, and Cohen (1997) suggested that the mechanism by which this method of job search leads to positive outcomes is at least partially due to the job seekers receiving accurate and detailed information about the new jobs. Overall, these studies suggest that networking may represent an effective method of job search from both the individual and organizational perspectives. To date, however, little is known about the individual-differences variables associated with the use of networking in job search and about the outcomes of networking in conjunction with other major determinants of reemployment (e.g., demographic factors, occupation and industry, and general job-search intensity). The following sections (a) further delineate and describe networking as a job-search technique, (b) outline two individual-differences variables (personality and networking comfort) that are assessed as predictors of networking intensity, and (c) summarize the reemployment outcomes (exhaustion of unemployment insurance, reemployment, and reemployment satisfaction) that are the focus of this study. Networking as a Job-Search Method Almost all individuals are members of social groups composed of family members, friends, and acquaintances. These unique social structures and relationships can be considered as different networks, with each individual comprising an element of a particular network, tied either directly or indirectly to other members of that network (Sprengers, Tazelaar, & Flap, 1988). Relationships existing between individuals allow for the transfer, or flow, of different kinds of information throughout the network. In the context of job search, Lowstuter and Robertson (1995) have defined networking as individual actions directed toward contacting friends, acquaintances, and other people to whom the job seeker has been referred for the main purpose of getting
information, leads, or advice on getting a job. Beatty (1988) noted that the networking process is a bit like cell division. In cell division, each parent cell divides into two cells. These cells then also divide into two more (total of four cells). Each of the resultant four cells then divide into two (total eight cells), and so on, as this process continues to explode at a rapid rate. It is also like the familiar chain letter, where each person copies the letter and gives it to five friends, who copy it and give it to five more friends, etc. The outcome is a geometric progression where the number of contacts increase at an ever increasing rate. (p. 168) Reid (1972) and Stevens et al. (1997) noted that some individuals do not use networking during their job search. Of those individuals who do use networking, many do not use the technique to its full potential (Stevens et al., 1997). Practitioner sources have noted that a systematic and complete use of the networking technique involves the unemployed person (a) creating a complete list of "Level 1" networking contacts, composed of friends, family, relatives, and people known from other realms of one's life (such as past bosses, subordinates, colleagues, clients, barbers, dentists, pastors, priests, lawyers, bankers, etc.); (b) informing these people that he or she is unemployed and looking for work; (c) asking for job leads or referrals to others who might be able to help (e.g., securing a list of "Level 2" networking contacts); and (d) contacting and gathering information from the Level 2 networking contacts (Azrin & Besalel, 1982; Beatty, 1988; Krannich & Krannich, 1996; Moock, 1996). As is the case for all aspects of the job search, the individual must also be ready to communicate his or her skills and job interests, conduct effective follow-up as appropriate, and keep up-to-date records on networking contacts and results. In the industrial-organizational psychology literature, research investigating whether individuals have used networking in their job search has commonly used one-item measures, such as "How many times have you talked with friends or relatives about possible job leads?" Typically, these items have been combined with other job-search behaviors to assess overall job-search intensity (defined as the sum of the frequency of use of all job-search methods), and the studies have focused on overall job-search intensity rather than on networking use per se (cf. Ellis & Taylor, 1983; R. Kanfer & Hulin, 1985; Vinokur & Caplan, 1987; Wanberg et al., 1996). In contrast, sociological investigations of networking have examined the types of contacts that people make during their job searches and which of these contacts tend to be most useful (cf. Lin & Dumin, 1986; Lin, Ensel, & Vaughn, 1981; Montgomery, 1992). Integrating the sociological and psychological literatures, we operationalize networking in this study as networking intensity and consider both the scope of networking behaviors used in the job-search process and the frequency with which these behaviors occur. Predictors of Networking Intensity Leaders of job-search workshops urge job seekers to use networking, and bookstores are laden with books urging the job seeker to include networking as a part of the job search. To date, however, little is known about the extent to which individualdifferences variables are related to job seekers' use of networking. Accordingly, this study examined individual differences in dispositional tendencies (i.e., personality) and individual differences in
UNEMPLOYED JOB SEEKERS comfort using a networking procedure as potential predictors of networking intensity in the job search.
Personality
Research devoted to the development of a taxonomy of personality traits has identified five broad, core dimensions of personality (the Big Five), known as Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness (cf. Digman, 1990). Neuroticism refers to the extent to which an individual displays anxiety, anger, hostility, self-consciousness, impulsiveness, vulnerability, and depression. Extraversion (also called surgency) refers to the extent to which an individual is outgoing, active, and high-spirited. Individuals with high levels of extraversion prefer to be around people most of the time. Individuals with high levels of openness to experience typically display imagination, curiosity, and originality and are open-minded. Individuals with high levels of agreeableness tend to be courteous, flexible, trusting, good-natured, cooperative, forgiving, empathetic, softhearted, and tolerant. Finally, individuals with high levels of conscientiousness are dependable (e.g., careful, thorough, responsible, organized, efficient, and good at planning) and have a high will to achieve (e.g., high achievement orientation and perseverance). No research has empirically assessed the relationship between the Big Five personality factors (or personality in general) and individuals' use of networking as a job-search technique. Research has, however, supported the importance of the Big Five to the job-search process in general. For example, Schmit, Amel, and Ryan (1993) reported that individuals with lower levels of neuroticism and higher levels of extraversion, openness to experience, agreeableness, and conscientiousness reported higher levels of assertive job-search behavior. Assertive job-search behavior was operationalized by Schmit et al. as a broader construct than networking, including readily asking questions when interviewed and applying for jobs even when one does not have all of the qualifications. Wanberg et al. (1996) similarly found that conscientiousness was associated with job-search intensity. Finally, Caldwell and Burger (1998) found that extraversion, along with openness to experience and conscientiousness to a lesser extent, was correlated with college students' use of social sources (e.g., talking to others) as a means of preparing for job interviews. Overall, the related literature, albeit limited and not directly involving the use of networking in the job search, is supportive of testing links between each of the Big Five factors and networking intensity. On the basis of this literature, we proposed that individuals with higher extraversion (e.g., because they are outgoing and active), agreeableness (e.g., because they are trusting, cooperative, good-natured, and have warm relationships with others), conscientiousness (e.g., because they are responsible, motivated, and persistent), and openness to experience (e.g., because they are flexible and open to trying different techniques and methods) display higher levels of networking intensity during the job-search process than individuals with lower extraversion, agreeableness, conscientiousness, and openness to experience. In addition, we proposed that individuals lower in neuroticism may be more likely to display high levels of networking intensity than individuals high in neuroticism, given that the latter group is more prone to selfconsciousness, hostility, and complaining when seeking out others
493
rather than using more constructive or problem-focused interactions (Watson & Hubbard, 1996). Hypothesis la: Higher levels of extraversion, agreeableness, consci-
entiousness, and openness to experience and lower levels of neuroticism are associated with higher levels of networking intensity. Of these five factors, however, we proposed that extraversion is especially portentous of an individual's frequency and scope of networking during the job search. The justification behind this proposition is twofold. First, a given level of networking intensity is more easily sustained the larger the number and size of one's networks or social capital (cf. Sprengers et al., 1988). Extraverts, individuals who by definition prefer the company of others and who are more frank and talkative (Mount & Barrick, 1995), can be expected to belong to larger and more diverse networks than introverts and thus will likely have a greater number of potential contacts (cf. Henderson, 1981; A. Kanfer & Tanaka, 1993; Russell, Booth, Reed, & Laughlin, 1997; Sarason, Levine, Basham, & Sarason, 1983). Meyer and Shadle (1994) noted that on the basis of experience in the outplacement field, individuals who are isolates (Wasserman & Faust, 1994) tend to have the most difficulty using networking methods. Second, given the availability of friends, family, and other potential contacts, one must consider how likely job seekers are to contact these individuals in their job search, as opposed to relying on other job-search methods. The research literature is suggestive of the relative importance of extraversion in comparison to the other Big Five traits in the engagement of behavior analogous to networking. For example, of the five personality traits, Watson and Hubbard (1996) found that extraversion was comparatively the most highly associated with coping with stressful events through seeking emotional and instrumental support from others. In a study of graduating college seniors, Caldwell and Burger (1998) conducted a canonical correlation between the Big Five personality variables and two measures of interview preparation: social preparation (e.g., talking to faculty, friends, or people in similar jobs) and background preparation (e.g., reading the material the company provided or looking for information about the company in magazines). Their canonical analysis identified two significant roots. The first root was depicted by social preparation and extraversion and, to a lesser extent, openness to experience. The second root was defined primarily by background preparation, high conscientiousness, and low neuroticism. Overall, the analysis highlighted a tendency for individuals who were extraverted (and, to a lesser extent, open to experience) to prepare for job interviews by drawing on and developing social contacts. Although conscientious individuals also used social contacts, these individuals had a slightly higher tendency to prepare for interviews by doing background research and reading. Extraversion includes in its facets not only sociability but also assertiveness and energy (Mount & Barrick, 1995), facets particularly conducive to a process that requires going beyond more passive methods of job search and information collecting. On the basis of this literature, we proposed to examine the following hypothesis: Hypothesis lb: Extraversion is more highly associated with network-
ing intensity than are the other Big Five dimensions.
494
WANBERG, KANFER, AND BANAS
Networking Comfort Azrin and Besalel (1982) discussed attitudinal barriers that may lead individuals to avoid using networking techniques. In general, it appears that individuals often avoid networking because they are uncomfortable with the idea or do not want to impose on a friendship. For example, Azrin and Besalel noted that some individuals may avoid networking because (a) they are embarrassed about being unemployed and may wish to avoid discussing it with others, or (b) they feel uncomfortable about asking for help or feel that networking is imposing on others. A comment by a participant in a qualitative study of graduating college students engaged in job search (Stevens et al., 1997) underscores this point: "I still am uncomfortable with the idea of calling people out of the blue to ask them questions that I really don't care to know the answer to, just so I can find out if there are any jobs available" (p. 22). Lowstuter and Robertson (1995, p. 54) similarly devoted a section of their book on networking to debunking the idea "I can't ask anyone for help. I don't do charity." In this study, we conceptualized networking comfort as a procedure-specific constellation of evaluative beliefs that portray an individual's attitudes toward using networking as a job-search method (see the Appendix for the items; further details are given in the Method section). Given that few persons would argue that networking is not a useful job-search method, we sought to differentiate individuals in terms of potential apprehensions held about using networking methods rather than on the basis of their global evaluations of the method per se. Likewise, we recognized that most individuals possess the minimal competencies necessary for executing specific networking behaviors (e.g., making a telephone call, talking to others about their job search) and did not seek to specifically assess personal judgments about whether one felt efficacious about being able to engage in the behavior. Instead, we examined the extent that individual differences in networking intensity stem from general discomfort about the use of networking because of factors such as apprehension about straining friendships or negative affect associated with perceiving oneself as looking bad to others. By defining attitudes toward networking in terms of comfort with a set of procedures, rather than tendencies to engage in social interaction in general, we distinguished the proximal construct (comfort with networking) from the related but distal construct (extraversion). Consistent with social-cognitive and motivational conceptions of the differential predictive validity of proximal and distal constructs (see, e.g., R. Kanfer, 1992), we expected that networking comfort would exert unique influence on networking intensity above that exerted by individual differences in extraversion. Specifically, we proposed the following:
Hypothesis 2a: Individual differences in comfort with networking are positively associated with networking intensity during job search. Hypothesis 2b: Individual differences in comfort with networking provide incremental predictive validity, beyond that of individual differences in extraversion, for networking intensity during job search. C o n s e q u e n c e s of Networking Intensity: Reemployment Outcomes The ultimate goal of networking during the job search is to find a job (ideally, a satisfactory job). On the basis of the empirical
(e.g., Granovetter, 1995) and practitioner (e.g., Krannich & Krannich, 1996) literature supporting the effectiveness of networking as a method of finding work, we proposed that higher levels of networking intensity are associated with lower levels of unemployment insurance exhaustion and increased probability and speed of reemployment. Although unemployment insurance exhaustion is correlated with reemployment probability and speed, we added this variable to our outcomes, as it is one closely attended to by state reemployment services because of the considerable financial resources devoted to unemployment benefits each year (cf. Minnesota Department of Economic Security [DES], 1999). Our specific hypothesis was as follows:
Hypothesis 3: Networking intensity is associated with lower levels of exhaustion of unemployment insurance benefits and positively predicts reemployment and reemployment speed. To provide a more complete account of the job search, however, we examined the influence of networking intensity along with general job-search intensity, defined as the frequency and scope of engagement in job-search behaviors such as reading the help wanted and classified ads, using the Internet to locate job openings, designing a good rtsumt, filling out a job application, and visiting an employment agency (typical job-search behaviors that are not directly within the networking-specific domain of contacting acquaintances or friends of friends). Previous research documents the usefulness of general job-search intensity in finding a job and the fact that individuals find jobs through both networking and nonnetworking procedures. We wished to assess the incremental prediction of networking behaviors beyond general job-search intensity. For example, for individuals already working hard with other methods, what additional advantages may be realized from employing networking strategies? We predicted that networking intensity would be predictive of our outcomes over and beyond general job-search intensity.
Hypothesis 4: Networking intensity negatively predicts lower levels of exhaustion of unemployment insurance benefits and positively predicts higher reemployment and reemployment speed after general job-search intensity has been accounted for. The inclusion of reemployment satisfaction in this study stems from our belief that in a strong economy, satisfaction with employment gained after a period of unemployment may be a more important criterion than simple reemployment and from research (reviewed earlier) suggesting a positive relationship between networking and satisfaction with a new job (e.g., Granovetter, 1995; Taylor, 1994). We used two indicators of reemployment satisfaction in this study--job satisfaction and intention to turnover--to test the following hypothesis:
Hypothesis 5: Individuals who report finding their jobs through networking report higher levels of job satisfaction and lower intentions to leave the organization than individuals who find their jobs through other methods. Method
Participants The participants in this study were unemployed at Time 1 and were recruited from 33 job service sites across the state of Minnesota. The participants were unemployment insurance recipients, permanently sepa-
UNEMPLOYED JOB SEEKERS rated from their last employer and actively seeking work, who were attending a half hour orientation that described the services provided by the state's DES for unemployed individuals. The participants had been mandated to attend the job service orientation through a program targeting individuals with characteristics associated with historically longer periods on unemployment insurance. All unemployment insurance recipients in Minnesota receive profiling quotients based on their occupation, industry, education level, and number of employers in the past 2 years. Individuals who are targeted to attend the orientation are more likely to have (a) professional and managerial, clerical and sales, processing, and benchwork occupations; (b) occupations located in the manufacturing, public administration, or finance-insurance-real estate industries; and (c) higher education levels than individuals who are not targeted. Protected class variables such as age, gender, and race are not used in the profilIng process. Despite the targeting procedure used, the orientation attendees were from a diverse set of occupations and industries and had a range of education levels (although, as is described later, the sample was more highly composed of professionals, individuals of higher education, and individuals from manufacturing industries than the population of this state's unemployment insurance claimants). During the 7-week period in 1997 that data were collected, 664 individuals attended the mandatory orientations. At the end of each orientation, individuals were asked if they were willing to complete a survey. After completing the survey, individuals folded it and placed it in an envelope for confidentiality purposes. Some individuals had appointments following the orientation and were not able to complete the surveys at the end of the orientation. These individuals were asked to return their surveys in a preaddressed postage-reply envelope. A total of 527 completed surveys were returned, for a response rate of 79%. Individuals were also asked to sign a release so that the researchers could obtain reemployment data from the DES. Releases were signed by 478 of the 527 individuals completing surveys (91%). The study focuses primarily on this sample of 478 individuals. Of the 478 study participants, 233 were men and 245 were women. Individuals ranged in age from 19 to 67 years (M = 40.5, SD = 10.9). The ethnicity of the respondents was 92% White, 3 0 Hispanic, 2% African American, 1% Native American, I% Asian American, and 1% other. The mean education level was 13.6 years (SD = 2.1). The mean income level according to wages reported by employers to the DES in 1998 was $28,474 (SD = $20,070). At Time 1 of this study the participants had been unemployed for a mean of 61 days and a mode of 30 days (SD = 66.2). Nine months later (at Time 2 of our study), 138 (29%) were still unemployed, whereas 340 (71%) had become reemployed. DES data indicated that 200 individuals (42%) had exhausted their allotted levels of unemployment insurance. Our sample included a higher proportion of individuals from professional, technical, and managerial occupations than the general population of Minnesota unemployment insurance recipients (45% in our sample vs. 18% in the general population of unemployment insurance recipients). However, all of the U.S. Department of Labor occupational categories were represented, including clerical and sales (n = 104), service (n = 34), agricultural and fishery (n = 6), processing (n = 15), machine trades (n = 12), bench work (n = 25), structural work (n = 20), and miscellaneous (n = 46). Our sample also included a higher proportion of participants from the manufacturing industry than the general population of Minnesota unemployment insurance recipients (31% in our sample vs. 24% in the general population of unemployment insurance recipients). However, all standard industry categories were represented, including agriculture, forestry, and fishery (n = 20); construction, mining, and quarrying (n = 21); transportation, communication, electric, gas, sanitary, and post office (n = 18); wholesale and retail Wade (n = 70); finance, insurance, and real estate (n = 33); services, hotel, repair, and recreation (n = 71); health, legal, education, engineering, and management (n = 82); and public administration (n = 14). Finally, our sample was more highly educated than the general sample of Minnesota unemployment insurance recipients. Whereas
495
28% of our sample had over 15 years of education, 11% of the general sample of unemployment insurance recipients in this state had over 15 years of education in calendar year 1997 (Minnesota DES, 1998). Data regarding benefits exhaustion, reemployment status, and date of reemployment for our sample of 478 individuals were obtained from DES data 9 months after Time 1. The 9-month follow-up was gauged to be sufficient to allow a majority of the respondents to become reemployed (the mean length of unemployment for the general unemployment insurance population in this state was 12.9 weeks in 1997, and individuals had already been unemployed prior to the Time 1 data collection). In addition to the agency data we received from the DES, a supplemental follow-up survey was also sent to the participants. The follow-up survey was designed to elicit self-report information from reemployed individuals on how they learned about their new jobs, their job satisfaction, and their intentions to quit their new jobs. A U.S. $1 silver certificate was included with each survey as an incentive to respond and as a token of appreciation. A total of 273 of these follow-up surveys (57%) were returned. Of the 273 participants wlao returned surveys, 192 (70%) were reemployed and thus provided the requested reemployment information.
Time 1 Measures Personality. The five personality domains (i.e., Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscientiousness) were measured with the NEO Five-Factor Inventory (NEO-FFI), a 60-item version of the 240-item Revised NEO Personality Inventory (Costa & McCrae, 1992). Each domain is measured by a 12-item scale. Individuals respond by using 5-point scales ranging from 1 (strongly disagree) to 5 (strongly agree). Costa and McCrae reported evidence supporting the internal consistency and validity of these scales. Three-month test-retest reliability of the NEO-FFI scales is estimated to range from .79 to .83 (Costa & McCrae, 1992). Alphas for the five scales in this study ranged from .72 to .85. Networking comfort. An eight-item scale was written to assess comfort level with the idea of networking (see the Appendix for specific items). Networking comfort was conceptualized as a procedure-specific constellation of proximal evaluative beliefs that depict an individual's attitude toward using networking as a job-search method. The items were written both to generally address the issue of comfort (e.g., "I am comfortable asking my friends for advice regarding my job search") and to specifically incorporate two barriers to networking discussed by Azrin and Besalel (1982). One barrier is embarrassment about being unemployed and not wishing to discuss the job search with friends, family, or friends of friends. A second barrier is discomfort with asking for help or feeling that networking is imposing on others. Items were completed by using 5-point Likert scales ranging from 1 (strongly disagree) to 5 (strongly agree). Exploratory factor analysis demonstrated that the eight items loaded on one factor. Confirmatory factor analysis results further supported this onefactor model (e.g., goodness-of-fit index, or GFI, = 0.95, comparative fit index, or CFI, = 0.91), with all factor loadings significant. The internal consistency reliability of the networking comfort scale was 0.79. Confirmatory factor analysis was also used to assess the distinctiveness of the networking comfort and networking intensity constructs. These analyses indicated that a two-factor model, in which networking comfort and networking intensity were specified as distinct factors, provided better fit indices (GFI = 0.89; CFI = 0.88) than did a one-factor model (GFI = 0.77; CFI = 0.75). Networking intensity. For purposes of this study, networking intensity was defined as the frequency and thoroughness of using networking in the job search (e.g., the frequency and thoroughness of contacting other people to get information, leads, or advice about job opportunities and the jobsearch process; Lowstuter & Robertson, 1995). Nine items were used to assess this construct (see the Appendix for specific items). Although we used Lowstuter and Robertson's (1995) definition of networking to anchor our item development (e.g., items inquired about contacting other people to
496
WANBERG, KANFER, AND BANAS
get information, leads, and advice), the items also necessarily reflect the multiple levels of networking, including creating a list of networking contacts, informing those contacts that one is looking for work, and asking for job leads or referrals to others (e.g., friends of friends) who might be able to help (Azrin & Besalel, 1982; Beatty, 1988; Krannich & Krannich, 1996; Moock, 1996). Six items inquired about the number of times individuals had engaged in various networking behaviors (the scale ranged from 1--never or zero times--to 5--very often, at least ten times), and three items inquired about the thoroughness of use of the networking process (the scale ranged from 1--I have not done this--to 5--1 have done a thorough job of this) over the course of a 2-week period. Seven of the networking intensity items were developed for this study, whereas the remaining two items ("Talked with friends or relatives about possible job leads"; "Spoke with previous employers or business acquaintances about their knowing of potential job leads") were derived from Blau's (1994) overall job-search intensity measure. We chose the last 2 weeks as a time frame for two reasons. Our first reason was psychometric in nature. We felt that by limiting the time frame, we could improve the accuracy of participants' recollection about the number of times they had engaged in the different job-search behaviors. Our second reason was that the 2-week time frame was tied to the nature of our sample. Given that many of our participants were recently unemployed, we felt our assessment of search behavior should correspondingly be for a time period of less than a month. Confirmatory factor analysis demonstrated good fit of the networking intensity items in a one-factor model (e.g., GFI = 0.89; CFI = 0.90; all factor loadings significant). The internal consistency of the networking intensity scale was 0.89. Confirmatory factor analysis was also used to assess the distinctiveness of networking intensity and general job-search intensity. Results of these analyses indicated that a two-factor model, in which networking intensity and general job-search intensity were specified as two factors, provided better fit indices (GFI = 0.88; CFI = 0.86) than did a one-factor model (GFI = 0.82; CFI = 0.77). General job-search intensity. General job-search intensity was defined as the frequency and scope of job-search activity without specific reference to networking and also without reference to the success of the search activity (e.g., having a job interview). Seven items were used to assess this construct (see the Appendix). The first six items listed in the Appendix were developed by Blau (1994). The seventh item ("Used the Worldwide Web or other computer services to locate job openings.") was added to capture the computer-based job-search activity that has become more common in the time since Blau's scale was written. Response options ranged from 1 (never or zero times) to 5 (very often, at least ten times). Research by Blau with samples of employed job seekers has shown support for scoring job-search intensity scales in terms of preparatory job-search behavior (e.g., reading want ads or preparing a resume) and active jobsearch behavior (e.g., sending out resumes). However, confirmatory factor analysis demonstrated that the items that were used in this study fit best to a one-factor model (e.g., GFI = 0.95; CFI = 0.87), rather than to a two-factor model with items categorized as preparatory or active (e.g., GFI = 0.88; CFI = 0.74). It is possible that the preparatory-active search dichotomy is more appropriate for an employed sample than for an unemployed sample. When individuals are unemployed, there is a greater blurring between preparing for a job search and actively engaging in a job search. For example, unemployed job seekers who are motivated likely check the want ads regularly. Their doing so is actually part of the active job search rather than a preparatory behavior. An employed person, in contrast, may think about (prepare for) the job search for some time by looking at want ads before actively engaging in a search and sending out resumes or contacting employers. Following this logic and the confirmatory factor analysis results, we scored the general job-search intensity scale by summing all of the items into one total. The internal consistency reliability of this scale was 0.71.
Time 2 Measures Benefits exhaustion, reemployment, and reemployment speed. Benefits exhaustion, reemployment status, and date of reemployment were obtained from state DES data 9 months after Time 1. Benefits exhaustion refers to whether an individual had reduced his or her unemployment insurance fund to $1 or less (1 = did exhaust; 0 = did not exhaust). Individuals are potentially eligible for a maximum of 26 weeks of state-provided unemployment insurance, but the time to exhaust their unemployment insurance may exceed 26 weeks if a claim is not filed every week. Reemployment status refers to whether or not the individual was reemployed 9 months after the initial survey (1 = yes; 0 = no). Reemployment status is reported to the DES through a variety of means, including employer's quarterly wage detail reports, new employee reports, and individual self-reporting. Speed of reemployment was operationalized as the total number of days an individual was unemployed. Whereas benefit exhaustion and reemployment status data were available for all 478 of our participants, speed of reemployment data were available for only 278 of the 340 reemployed individuals (82%). Method of reemployment. Reemployed individuals were asked in the supplemental follow-up survey to report how they learned about the job they now held. Response options included checking want ads and advertisements; searching the Internet; talking to friends, family, and previous coworkers; networking; attending a job fair; going to a workforce center, job service, or other employment agency, and other. Job satisfaction and intentions to turnover. Job satisfaction (e.g., "All in all, I am satisfied with my job") and intention to turnover (e.g., "I often think about quitting") were assessed by using two 3-item scales from the Michigan Organizational Assessment Questionnaire (Cammann, Fichman, Jenkins, & Klesh, 1983) that use 1- to 7-point responses. These measures were completed by reemployed individuals in the supplemental follow-up survey. The internal consistency reliability values of these two measures were .93 (job satisfaction) and .89 (intention to turnover). Control Variables Gender, race, education, age, weekly benefit amount, and days unemployed at Time 1 were used as control variables. Gender was coded 0 = male and I = female. Race was coded 0 = minority and 1 = White. For education, individuals were asked to circle the highest year of education they had completed (1 to 17 or more). Age was assessed with a fill-in-theblank question and verified with DES data. Weekly benefit amount (the amount in dollars each individual was qualified to receive per week for unemployment insurance) was obtained from DES records. Days unemployed at Time 1 was assessed with an item in the survey that asked "How long have you been unemployed? Please estimate as accurately as possible." Blanks were left for number of months, weeks, and days the individual had been unemployed, and a calendar was provided to help participants' recollection. The time unemployed was coded into total days unemployed at Time 1. These control variables have been implicated as potentially important in job-search and unemployment research. According to Leana and Feldman (1992), women, minorities, individuals with lower education, and individuals who are older may face longer periods of unemployment and may cope differently with job loss. In regard to weekly benefit amount, economic research has shown that receipt of unemployment benefits lowers reemployment speed (Barron & Mellow, 1981) and that higher levels of unemployment benefits decrease search intensity (Barron & Gilley, 1979). Number of days unemployed may also play an important role in the job-search process. As unemployment length increases, individuals are more likely to have higher levels of financial strain and anxiety (Wart & Jackson, 1984), and they may change their search intensity or patterns (Barber, Daly, Giannantonio, & Phillips, 1994). Industry and occupational codes were initially used in our regressions as control variables. Seventeen dummy variables were used to represent 9 major occupational codes and 10 major industry codes. None of these
UNEMPLOYED JOB SEEKERS dummy variables was significant in the prediction of benefit exhaustion, reemployment status, or reemployment speed. In the prediction of networking intensity, one industry code variable was significant. Individuals in the manufacturing of computer, electrical, and transportation equipment industry reported less networking than individuals in the dummy comparison group of agriculture, forestry, and fishery (p < .05). In light of the general nonsignificance of the occupation and industry variables and for consequential parsimony considerations, the decision was made to exclude these variables from further consideration in the analyses. It is important to note that their exclusion did not change any of the results (e.g., reported significance of other variables) reported in this article.
Analyses Mean substitution (Roth, 1994) was used to calculate scale scores for a small percentage of participants who did not complete one or two items on a scale. Mean substitution was not used on categorical items (such as how reemployed individuals found their new jobs). Instead, missing data were left missing for these items. The largest number of mean replacements for any one item was nine, representing 2% of the sample. Hierarchical multiple regression was used for the prediction of networking intensity. The control variables were entered in Step 1, the personality variables in Step 2, and networking comfort in Step 3 to allow for an assessment of the relative and incremental contribution of the predictors. Hierarchical logistic regression was used for the prediction of unemployment insurance exhaustion and reemployment at Time 2 of this study, because these variables are dichotomous (1 = yes, 0 = no). The logistic regression coefficients presented are odds ratios (labeled as Exp B in SPSS; Norusis, 1994). Odds ratios are not standardized and thus cannot be used to evaluate the relative strength of the predictors. Instead, an odds ratio that is significant and greater than 1 indicates that the odds of the outcome variable increase when the predictor increases. An odds ratio that is significant and less than 1 indicates that the odds of the outcome variable decrease when the predictor increases (Menard, 1995). Model chi-square values are provided for each equation computed and are analogous to the multivariate F tests used in linear regression (DeMaris, 1992). Classification accuracy percentages are provided as supplemental information, although it should be noted that these estimates are sensitive to the observed percentage of individuals in the predicted groups (Hosmer & Lemeshow, 1989). Hierarchical multiple regression was used for the prediction of job satisfaction and turnover intention. Results Table 1 reports the means, standard deviations, coefficient alphas, and correlations among the variables used in this study. The correlation between the two job-search variables (networking intensity and general job-search intensity) was .49, indicating that expending effort in networking activities is associated with greater uses of other job-search methods. The highest predictor-predictor correlations were between Agreeableness and Conscientiousness (r = .44), Extraversion and Neuroticism (r = -.45), and Conscientiousness and Neuroticism (r = -.43).
Prediction of Networking Intensity Hypotheses la and lb address the relationship between the personality scales and networking intensity. In support of Hypothesis la, each of the Big Five personality variables was significantly correlated with networking intensity in the direction expected: Neuroticism (-.19), Extraversion (.34), Openness (.13), Agreeableness (.10), and Conscientiousness (.21; see Table 1). Table 2 portrays multivariate results. When networking intensity was regressed on all five of the personality variables, only Extraversion
497
and Conscientiousness were significantly related to networking intensity. Hypothesis lb suggests that Extraversion is more highly associated with networking intensity than the other personality variables are. Showing partial support for this hypothesis, a test of equivalence of betas (Neter, Wasserman, & Kuttner, 1990) showed that the Extraversion beta was significantly larger than all of the others except for Conscientiousness. Hypotheses 2a and 2b propose that higher levels of networking comfort are associated with higher networking intensity and that networking comfort is predictive of networking intensity above and beyond the effects of personality. As shown in Table 2, these hypotheses were supported. The addition of networking comfort in Step 3 of the equation for networking intensity was associated with an increment in R 2 of .14 (p < .01). It is also interesting to note that when networking comfort was entered into the equation, Conscientiousness was no longer significantly associated with networking intensity. A final observation is that none of the control variables were significant in the networking intensity regression equations. For comparison purposes, Table 2 shows the regression of general job-search intensity on the same variables. Step 2 of the equation shows that higher levels of Extraversion and Conscientiousness were related to higher levels of general job-search intensity. Although these variables remained significant in Step 3 when networking comfort was added to the equation, networking comfort itself was not significant in predicting general job-search intensity. Of the control variables, only education was (positively) associated with general job-search intensity.
Prediction of Reemployment and Reemployment Quality Hypothesis 3 suggests that networking intensity predicts lower unemployment insurance exhaustion, increased probability of reemployment, and faster reemployment. The results for this hypothesis are shown under the columns marked H3 in Table 3. Hypothesis 4 predicts that networking intensity is predictive of the reemployment outcomes above and beyond general job-search intensity (see columns marked H4 in Table 3). Logistic regression was used to generate the results appearing for the columns marked Unemployment insurance benefit exhaustion and Reemployment. The values appearing for these analyses are logistic regression odds ratios. An odds ratio that is significant and greater than 1 indicates that the odds of the outcome variable increase when the predictor increases. An odds ratio that is significant and less than 1 indicates that the odds of the outcome variable decrease when the predictor increases (Menard, 1995). Supporting Hypothesis 3, the results indicate that, without general job-search intensity in the equation, greater networking intensity is related to lower likelihood of exhausting unemployment benefits and to a greater likelihood of reemployment at Time 2 of this study. The logistic regression odds ratio of 0.97 predicting benefit exhaustion indicates that a one-unit increase in networking intensity (e.g., contacting or talking to three more people) decreases the odds of benefit exhaustion by 3%. The odds ratio of 1.03 in the equation predicting reemployment indicates that a one-unit increase in networking intensity increases the odds of reemployment by 3%. Hierarchical multiple regression was used to predict total days unemployed. Contrary to expectations, networking intensity was not related to speed of reemployment.
498
WANBERG,
KANFER,
AND
BANAS
I
I
r-'.-
I
,
II I
I
.~ I[
¢¢
o~ O~
I
I
_,7. ,.~x5 I
I oc
I
I
~ ~*.--- 8 8 I
I
I
t
I
I
I
O
~8
I
I
I
I
I
I
I
I O
I
-~o~;.
~ .
I
I
I
I
I
k.
I
I
I
tr) oq. ¢~. ,---~. ' ~
t~
I
I
I
I
I
I
I
I
,..-. O", ¢,n r ' - ~:~ t'-,i e ~ t'~ ,--~ ~,-~ ~t-~
8o
I I
I
e q ,,--~ ,,~.
O
UNEMPLOYED JOB SEEKERS
499
Table 2 Unemployed Individuals at Time 1: Predictors of Networking Intensity and General Job-Search Intensity Networking intensity (/3) Predictor Control variable Gender (0 = male, 1 = female) Age Race (0 = minority, 1 = White) Education Weekly benefit amount Time 1 days unemployed Predictor variable Neuroticism Extraversion Openness to Experience Agreeableness Conscientiousness Networking comfort R Adjusted R 2 ~kR2
Step 1 -.01 - .07 .02 .10 .03 -.05
.14 .01 .02
Step 2
General job-search intensity (/3) Step 3
-.03 - .05 .02 .04 .01 -.01
-.05 - .02 .00 .06 -.03 .02
- .01 .28** .06 - .02 .13"
.03 .20** .04 - .04 .09 .40**
.37** .12"* .12"*
.53** .26** .14'*
Step 1
Step 2
Step 3
-.01 -.09 -.01 .20** -.05 .04
-.05 - .08 -.02 .17"* -.06 .07
-.05 - .08 -.02 .17"* -.06 .07
.05 .13" .07 .00 .20**
.05 .12" .07 .00 .19"* .03
.32** .08** .06**
.32** .08** .00
.20** .03** .04**
Note. n = 478. The values in the table are standardized beta weights. Step 1 portrays results with only the control variables in the equation. Step 2 portrays the results with the addition of the five personality variables. Step 3 portrays the results with the addition of networking comfort. * p < . 0 5 . **p .05. Similarly, individuals who found their jobs through networking did not report lower intentions to turnover (M = 8.5, SD = 5.1) than individuals who did not find their jobs through networking (M = 9.3, SD = 5.6), t(180) -- 0.92, p > .05. Discussion Popular bookstores are laden with books to advise individuals about their job searches. Book rifles shout Network Your Way to Your Next Job: Fast (Lowstuter & Robertson, 1995) and Dynamite Networking for Dynamite Jobs (Krannich & Krannich, 1996). Much, however, is left to be learned about how to encourage networking behavior among individuals and about the effectiveness of this method in comparison with other methods of job search. This study contributes to the existing literature on jobsearch behavior by assessing both the predictors and outcomes of networking behavior. Although all of the Big Five variables and networking comfort were significantly correlated with networking intensity, Extraversion, Conscientiousness, and networking comfort were highlighted in the results because of their significance in the multivariate
500
WANBERG, KANFER, AND BANAS Table 3 Networking and Job-Search Intensity as Predictors o f Reemployment Outcomes
Exp B UI benefit exhaustion Predictor Gender (0 = male, 1 = female) Age Race (0 = minority, 1 = White) Education Weekly benefit amount Time 1 days unemployed General job-search intensity Networking intensity % classification accuracy Logistic regression model chi-square Chi-square improvement R Adjusted R2 ~1~2a
Reemployment
H3
H4
H3
H4
1.05 1.02' 0.64 0.93 1.00 1.00
1.04 1.02' 0.61 0.96 1.00 1.00 0.92** 1.00
1.09 0.98* 2.06* 1.00 1.00 1.00
1.10 0.98* 2.13" 0.98 1.00 1.00 1.07"* 1.01
63.8 30.6 (8)** .00 (1)
70.3 15.9 (7)*
0.97* 60.7 15.5 (7)*
1.03"
Total days unemployed (~) H3
H4
.02 .02 .11' .10' -.03 -.05 - .08 - .05 -.01 -.01 .25** .24** -.16" .03 .10
71.8 23.8 (8)** .27 (1) .29** .06**
.32** .08** .01
The first four columns portray logistic regression equations predicting unemployment insurance (UI) benefits exhaustion and reemployment (n = 478). The coefficients in these equations (Exp B) are logistic regression odds ratios. The last two columns portray multiple regression analyses predicting reemployment speed for those individuals who were reemployed at Time 2 (n = 278 because of missing data). The coefficients in these last two columns are standardized beta weights. Degrees of freedom for chi-squares are in parentheses. Empty cells indicate that general job-search intensity was not included as a variable in that model. For UI benefit exhaustion and reemployment, 1 = yes and 0 = no. The equations marked H3 assess Hypothesis 3, that networking intensity is associated with the reemployment outcomes. Equations marked H4 assess Hypothesis 4, that networking intensity predicts the reemployment outcomes after general job-search intensity has been accounted for. a The chi-square improvement and/~kR2 statistics indicate the change when the networking intensity variable is added to a model already containing job-search intensity. *p < .10 (marginally significant). *p < .05. **p < .01. Note.
analyses. Of these three variables, networking comfort is the most proximal (and least trait-based) variable and thus may arguably have the clearest implications for interventions with the unemployed. For example, our findings suggest that job-search professionals who wish to teach and encourage networking as a jobsearch tool should work on increasing networking comfort levels among their clients. Sociological research (e.g., Granovetter, 1973) has suggested that networking during the job search is usually most successful through individuals with whom the job seeker has weak ties rather than strong ties. For starters, job seekers must be comfortable with the idea of contacting their friends, family, and acquaintances and talking to them about their job search. Subsequently, they have to feel comfortable with the idea of contacting weaker ties or the friends of friends--individuals with access to different information. Individuals must be encouraged to open up to others and to put aside ideas that networking means bothering others or admitting failure. The associations between conscientiousness, extraversion, and networking intensity are also meaningful for job-search professionals. Individuals low in conscientiousness can be encouraged to set goals and to be more purposeful with their networking endeavors. Regarding the trait of extraversion, it is useful for these professionals to recognize that introverted individuals are at a disadvantage not only because they are disinclined to confide in or socialize with others but also because of the fact that they likely have smaller networks of acquaintances, or less social capital (Sprengers et al., 1988), than individuals who are more extro-
verted. Introverted job seekers might benefit from explicitly making a list of contacts, listing everyone they know, including people whom they do not know well but who may be able to help. For example, although research by Sprengers et al. (1988) showed that almost 80 job seekers of a sample of 242 reported they did not have any friends, all had relatives, and it is likely that many of them could have listed other potential contacts. Introverted individuals may also be encouraged to take part in networking job club meetings where they are encouraged to discuss their job search with others. Some networking clubs require that every member make an average of two referrals a week on the behalf of another member, and valiant attempts are made to make sure every member is receiving referrals (Baker, 1994). It is also possible that introverted individuals may need more encouragement than others to maintain an intensive job search across the board regardless of the methods used. Our findings, for example, indicate that extraversion is not only related to networking intensity but also to general job-search intensity. Clearly, the ultimate goal of networking during the job search is to find a job. Supporting the usefulness of networking as a jobsearch method, more than 36% of the individuals in our study who completed our follow-up mail survey indicated that they found their jobs through networking or through talking to friends, family, or previous coworkers. This figure is roughly comparable to others cited in the literature (cf. Granovetter, 1995). Our results also indicate that networking intensity is associated with lower levels of unemployment insurance benefit exhaustion and an increased
UNEMPLOYED JOB SEEKERS probability of reemployment, suggesting that individuals who use networking more often and more extensively have a relative advantage over individuals who use networking less often and less extensively. Yet, despite the clear usefulness of networking as a job-search method, our study provides an important limitation to the advantages rendered by this method. Specifically, when considered in the more complete context of individual differences in general job-search intensity, no evidence was found for the incremental contribution of networking intensity in the prediction of these outcome variables. These findings suggest that although networking is useful, claims should not be made that this method is superior to traditional job-search techniques. Consistent with this idea, Meyer and Shadle (1994) stressed that networking should be used as just one job-search tool and that it should not be used at the exclusion of others. Our findings even suggest that if a person is expending high effort in his or her job search through the use of other job-search methods, networking may not be necessary. We recommend, however, that job-search professionals continue to encourage networking as a job-search method, as networking does indeed lead individuals into new jobs. We also feel it is still valuable to hold workshops that focus on how to use networking in the job search. However, job-search professionals should communicate to individuals that networking is best used as a complement to other job-search methods. It is possible that networking intensity would add incremental validity above and beyond general job-search intensity in the context of more recessionary economic conditions, or at least under conditions where labor markets are less favorable for job seekers. At the current time, however, with low unemployment rates and abundant jobs (the unemployment rate in this state ranged from 2.5% to 3.5% during the 9 months these data were collected, below the nation's average), a person who is diligently using formal or traditional job-search methods such as want ads and other advertisements is likely to be successful without the frequent or thorough use of networking. Future research is needed to assess the generalizability of these results in less munificent economic times. Although our analyses incorporating occupational and industry data were not suggestive in this regard, it is also possible that networking is critical to specific types of jobs or in certain types of situations. A discussion of limitations of this study suggests other cautions regarding our conclusions, along with further research needs in this area. First, despite our decision to assess both networking and job-search intensity for the last 2 weeks (e.g., to improve accuracy of recall), it is possible that this short time frame was not sufficient to provide an accurate and complete portrayal of an individual's job-search patterns. Future studies may benefit from assessing job-search intensity at multiple time waves. Multiple time waves asking about job search in the last 2 weeks would more fully capture individual differences in job-search intensity while still being sensitive to concerns about accuracy of recall. It is possible that another explanation for the failure for networking intensity to predict reemployment success over and beyond general job-search intensity may have been that we did not fully tap networking frequency. Future research in this area should simultaneously attempt to more clearly identify the process by which job searchers contact others and acquire information and the extent to which networking intensity leads to the derivation of specific and detailed informa-
501
tion about jobs (e.g., a realistic job preview from appropriate contacts; Griffeth et al., 1997). For example, the results derived from this study regarding the relationship between networking intensity, job satisfaction, and intention to turnover are limited given that we did not assess the extent to which individuals collected realistic job preview information during their networking. It may be that there would have been a relationship between finding a job through networking (vs. another method) and job satisfaction if the extent to which job preview information had been obtained from contacts had been accounted for. Another missing component in this study was an assessment of networking quality on the part of the participants. The extent to which an individual is doing a lot of networking does not tell us whether or not he or she is doing it effectively. Networking professionals often note that it is possible to be too bold or pushy when networking. Similarly, there is a quality issue on the part of the contacts. It is possible that it is not the intensity of networking that matters but whom one has available to contact. Finally, future research should refine measures of the constructs assessed in this study. Although confirmatory factor analyses supported the dimensionality of the scales developed for use in this study (e.g., networking comfort and networking intensity), the fit indices for these scales indicated that they could be improved. The variable networking comfort would also especially benefit from research assessing its discriminant validity from related (but more general) concepts that have been examined in the literature (e.g., job-search self-efficacy; Caplan, Vinokur, Price, & van Ryn, 1989). Despite these limitations, this study makes important contributions in several areas. Although we were not able to collect time series data, our study was longitudinal in nature and extended previous research with a follow-up period of 9 months in comparison with what is commonly no more than 3 months in this literature. A second contribution lies in the large sample size with reemployment outcomes obtained through state DES records. Previous research on the topic of unemployment and job search has heavily relied on self-report survey data. We were able to report objective DES outcome data on 72% of our potential respondents, representing a much higher response rate than many previous studies in this area. Study results shed light on a topic (networking as a method of job search) that has received very little empirical attention in comparison with the weight that it has been given in the practitioner literature. Finally, our results extend a growing literature on the relationship between the Big Five and a host of work-related outcomes (cf. Tokar, Fischer, & Subich, 1998). In conclusion, this study contributes to the literature on unemployment and job search. Although research on the general topic of social networking and networking in the organizational environment is increasing (cf. Shah, 1998; Tjosvold, 1997), few studies have been devoted to networking in the job-search context. This study sheds insight into the types of individuals most likely to use networking in their job search. It further demonstrates that networking, although a useful and important job-search method, should not be encouraged at the expense of other job-search methods. Future research is needed to assess the generalizability of these findings to other economic or cultural contexts. References Allen, R. E., & Keaveny, T. J. (1980). The relative effectiveness of alternative job sources. Journal of Vocational Behavior, 16, 18-32.
502
WANBERG, KANFER, AND BANAS
Azrin, N. H., & Besaiel, V. B. (1982). Finding a job. Berkeley, CA: Ten Speed Press. Baker, W. E. (1994). Networking smart. New York: McGraw-Hill. Barber, A. E., Daly, C. L., Giannantonio, C. M., & Phillips, J. M. (1994). Job search activities: An examination of changes over time. Personnel Psychology, 47, 739-765. Barron, J. M., & Gilley, O. W. (1979). The effect of unemployment insurance on the search process. Industrial and Labor Relations Review, 32, 363-366. Barron, J. M., & Mellow, W. (1981). Changes in labor force status among the unemployed. Journal of Human Resources, 16, 427-441. Beatty, R. H. (1988). The complete job search book. New York: Wiley. Blau, G. (1990). Exploring the mediating mechanisms affecting the relationship of recruitment source to employee performance. Journal of Vocational Behavior, 37, 303-320. Blau, G. (1993). Further exploring the relationship between job search and voluntary individual turnover. Personnel Psychology, 46, 313-330. Blau, G. (1994). Testing a two-dimensional measure of job-search behavior. Organizational Behavior and Human Decision Processes, 59, 288312. Breaugh, J. A., & Mann, R. B. (1984). Recruiting ~ource effects: A test of two alternative explanations. Journal of Occupational Psychology, 57, 261-267. Bureau of Labor Statistics. (1998, July 24). Labor force statistics from the current population survey [On-line]. Available: http://146.142.4.24/cgibin/surveymost?lf Washington, DC: Author. Caldwell, D. F., & Burger, J. M. (1998). Personality characteristics of job applicants and success in screening interviews. Personnel Psychology, 51, 119-136. Cammann, C., Fichman, M., Jenkins, D., & Klesh, J. (1983). Assessing the attitudes and perceptions of organizational members. In S. E. Seashore, E. E. Lawler HI, P. H. Mirvis, & C. Cammann (Eds.), Assessing organizational change: A guide to methods, measures, and practices (pp. 71-138). New York: Wiley. Caplan, R. D., Vinokur, A. D., Price, R. H., & van Ryn, M. (1989). Job seeking, reemployment, and mental health: A randomized field experiment in coping with job loss. Journal of Applied Psychology, 74, 759-769. Costa, P. T., Jr., & McCrae, R. R. (1992). Revised NEO Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI): Professional manual. Odessa, FL: Psychological Assessment Resources. DeMaris, A. (1992). Logit modeling: Practical applications. Newbury Park, CA: Sage. Digman, J. M. (1990). Personality structure: Emergence of the five-factor model. Annual Review Of Psychology, 41, 417-440. Ellis, R. A., & Taylor, M. S. (1983). Role of self-esteem within the job search process. Journal of Applied Psychology, 68, 632-640. Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 78, 1360-1380. Granovetter, M. S. (1995). Getting a job (2nd ed.). Chicago: University of Chicago Press. Griffeth, R. W., Horn, P. W., Fink, L. S., & Cohen, D. J. (1997). Comparative tests of multivariate models of recruiting sources effects. Journal of Management, 23, 19-36. Henderson, S. (1981). Neurosis and the social environment. New York: Academic Press. Hosmer, D. W., & Lemeshow, S. (1989). Applied logistic regression. New York: Wiley. Kanfer, A., & Tanaka, J. S. (1993). Unraveling the web of personality judgements: The influence of social networks on personality assessment. Journal of Personality, 61, 711-738. Kanfer, R. (1992). Work motivation: New directions in theory and research. In C. L. Cooper & I. T. Robertson (Eds.), International review of industrial and organizational psychology (Vol. 7, pp. 1-53). New York: Wiley.
Kanfer, R., & Hulin, C. L. (1985). Individual differences in successful job searches following lay-off. Personnel Psychology, 38, 835-847. Krannich, C. R., & Krannich, R. L. (1996). Dynamite networking for dynamite jobs. Manassas Park, VA: Impact. Latham, V. M., & Leddy, P. M. (1987). Source of recruitment and employee attitudes: An analysis of job involvement, organizational commitment, and job satisfaction. Journal of Business & Psychology, 1, 230 -235. Leana, C. R., & Feldman, D. (1992). Coping with job loss. New York: Lexington Books. Lin, N., & Dumin, M. (1986). Access to occupations through social ties. Social Networks, 8, 365-385. Lin, N., Ensel, W. M., & Vaughn, J. C. (1981). Social resources and strength of ties: Structural factors in occupational status attainment. American Sociological Review, 46, 393-405. Lowstuter, C. C., & Robertson, D. P. (1995). Network your way to your next job: Fast. New York: McGraw-Hill. Menard, S. (1995). Applied logistic regression analysis. Thousand Oaks, CA: Sage. Meyer, J. L., & Shadle, C. C. (1994). The changing outplacement process: New methods and opportunities for transition management. Westport, CT: Quorum Books. Minnesota Department of Economic Security. (1998, July 30). Profile of Minnesota Reemployment Insurance Claimants Calendar Year 1997 IOn-line]. Available: http://www.des.state.mn.us/lmi/UI/ri-prof7.htm. St. Paul, MN: Author. Minnesota Department of Economic Security. (1999, August 4). Minnesota Reemployment Insurance Claims Calendar Year 1998 IOn-line]. Available: http://www.des.state.mn.us/lmi/Ul/ri-sum.htm. St. Paul, MN: Author. Montgomery, J. (1992). Job search and network composition: Implications of the strength-of-weak-ties hypothesis. American Sociological Review, 57, 586-596. Moock, R. T., Jr. (1996). Get that interview, t The i~dispensable guide for college grads. Hauppauge, NY: Barron's Educational Series. Mount, M. K., & Barrick, M. R. (1995). The Big Five personality dimensions: Implications for research and practice in human resources management. Research in Personnel and Human Resources Management, 13, 153-200. Neter, J., Wasserman, W., & Kuttner, M. H. (1990). Applied linear statistical models: Regression, analysis of variance, and experimental designs (3rd ed.). Homewood, IL: Irwin. Norusis, M. J. (1994). SPSS advanced statistics (Release 6.1) [Computer software manual]. Chicago: SPSS. Price, R. H., & Vinokur, A. D. (1995). Supporting career transitions in a time of organizational downsizing. In M. London (Ed.), Employees, careers, and job creation (pp. 191-209). San Francisco: Jossey-Bass. Reid, G. L. (1972). Job search and the effectiveness of job-finding methods. Industrial and Labor Relations Review, 25, 479-495. Rosenfeld, C. (1975). Job seeking methods used by American workers. Monthly Labor Review, 98, 39-42. Roth, P. L. (1994). Missing data: A conceptual review for applied psychologists. Personnel Psychology, 47, 537-560. Rowley, K. M., & Feather, N. T. (1987). The impact of unemployment in relation to age and length of unemployment. Journal of Occupational Psychology, 60, 323-332. Russell, D. W., Booth, B., Reed, D., & Laughlin, P. R. (1997). Personality, social networks, and perceived social support among alcoholics: A structural equation analysis. Journal of Personality, 65, 649692. Sarason, I. G., Levine, H. M., Basham, R. B., & Sarason, B. R. (1983). Assessing social support: The Social Support Questionnaire. Journal of Personality and Social Psychology, 44, 127-139. Schmit, M. J., Amel, E. L., & Ryan, A. M. (1993). Self-reported assertive
UNEMPLOYED JOB SEEKERS job-seeking behaviors of minimally educated job hunters. Personnel Psychology, 46, 105-124. Schwab, D. P., Rynes, S. L., & Aldag, R. J. (1987). Theories and research on job search and choice. Research in Personnel and Human Resources Management, 5, 129-166. Shah, P. P. (1998). Who are employees' social referents? Using a network perspective to determine referent others. Academy of Management Journal, 41, 249-268. Sprengers, M., Tazelaar, F., & Flap, H. D. (1988). Social resources, situational constraints, and re-employment. Netherlands Journal of Sociology, 24, 98-116. Stevens, C. K., Tirnauer, D. C., & Turban, D. B. (1997, August). A qualitative study of job seekers' decision processes. Paper presented at the meeting of the Academy of Management, Boston, MA. Swaroff, P. G., Barclay, L. A., & Bass, A. R. (1985). Recruiting sources: Another look. Journal of Applied Psychology, 70, 720-728. Taylor, G. S. (1994). The relationship between sources of new employees and attitudes toward the job. Journal of Social Psychology, 134, 99-110. Tjosvold, D. (1997). Networking by professionals to manage change:
503
Dentists' cooperation and competition to develop their business. Journal
of Organizational Behavior, 18, 745-752. Tokar, D. M., Fischer, A. R., & Subich, L. M. (1998). Personality and vocational behavior: A selective review of the literature: 1993-1997. Journal of Vocational Behavior, 53, 115-153. Vinokur, A., & Caplan, R. D. (1987). Attitudes and social support: Determinants of job-seeking behavior and well-being among the unemployed. Journal of Applied Social Psychology, 17, 1007-1024. Wanberg, C. R., Watt, J. D., & Rumsey, D. J. (1996). Individuals without jobs: An empirical study of job-seeking behavior and reemployment. Journal of Applied Psychology, 81, 76-87. Wart, P., & Jackson, P. (1984). Men without jobs: Some correlates of age and length of unemployment. Journal of Occupational Psychology, 57, 77-85. Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications. Cambridge, England: Cambridge University Press. Watson, D., & Hubbard, B. (1996). Adaptational style and dispositional structure: Coping in the context of the five-factor model. Journal of Personality, 64, 737-774.
Appendix Networking Comfort, Networking Intensity, and General Job-Search Intensity Items Networking C o m f o r t 1. I am comfortable asking my friends for advice regarding my job search. 2. I don't like to bother people about my job search because I know they are busy. (reverse scored) 3. I don't mind asking family members and relatives if they have any job leads for me. 4. I am comfortable asking previous coworkers or acquaintances for their assistance in my job search. 5. I don't like to ask people for job leads or advice because it puts them on the spot or imposes on them. (reverse scored) 6. I am comfortable following up with people once I have contacted them about my interest in finding a job. 7. I am embarrassed about being unemployed and don't like to talk about it with others. (reverse scored) 8. I don't like to call friends of friends about possible job openings. (reverse scored)
Networking Intensity How often have you done each of the following in the last two weeks? 1. Contacted people you know to ask for their advice or leads regarding your job search. 2. Called or visited someone just to get more information about a certain job or place to work. 3. Asked for a referral to someone who might have helpful information or advice about your career or industry. 4. Secured leads from contacts or acquaintances regarding a person to contact for information that would help you in your job search. 5. Talked with friends or relatives about possible job leads. 6. Spoke with previous employers or business acquaintances about their knowing of potential job leads.
To what extent have you done each of the following in the last two weeks? 1. I have made a list of all the people I know that might have job leads or ideas for me (e.g., business colleagues, friends, neighbors, relatives, high school or college contacts, fellow members of religious institutions, etc.). 2. I have told many people (e.g., people in my family, personal friends, neighbors, and acquaintances) that I am unemployed and have explained the type of job I am looking for. 3. I have asked people I know if they knew of someone who might have job leads or information for me. General Job-Search Intensity How many times have you done each of the following in the last two weeks? 1. Read the help wanted/classified ads in a newspaper, journal, or professional association. 2. Prepared/revised your resume. 3. Read a book or article about getting a job or changing jobs. 4. Sent your resumes to potential employers. 5. Filled out a job application. 6. Contacted an employment agency, executive search firm, or state employment service. 7. Used the Internet (WWW or Worldwide Web) or other computer services to locate job openings. Items 1-6 are from '"l'esting a Two-Dimensional Measure of Job-Search Behavior," by G. Blau, 1994, Organizational Behavior and Human Decision Processes, 59, p. 298. Copyright 1994 by Academic Press. Reprinted with permission. Received D e c e m b e r 17, 1998 Revision received August 10, 1999 Accepted August 23, 1999 •
