Faculty and Institutional Review Board Communication - The National ...

Communication Education Vol. 58, No. 4, October 2009, pp. 497515

Faculty and Institutional Review Board Communication Michael W. Kramer, Vernon D. Miller & Suraj Commuri

The researcherInstitutional Review Board (IRB) relationship is a critical element in a faculty member’s role in knowledge generation and dissemination. The nature of this relationship has considerable implications for data gathering in the research process, as well as in the mentoring and instruction of graduate and undergraduate students as researchers. This study identifies communicative elements in the researcherIRB relationship and their impact on researchers’ attitudes toward IRBs and compliance with IRB submission processes. Implications of the findings are discussed with respect to scholars’ interactions with the regulatory agency, conducting research, and the education and training of students. As one of their roles, faculty members at an increasing number of universities are involved in the generation of knowledge through research. Their success in publishing their research has important implications for their promotion and tenure. When human subjects are involved, these scholars must receive approval from their Institutional Review Boards (IRBs) to conduct that research. In addition to communicating with IRBs to seek approval of their research plans when conducting their own research, the researchers model and teach their attitudes and behaviors to others when they lead graduate student research teams, supervise theses and dissertations, and direct graduate and undergraduate class research projects. As a result, the quality of communication interaction with the IRB affects researchers’ attitudes and behaviors, and the subsequent communication to students about IRBs helps shape their attitudes and behaviors as an important part of the educational process at many colleges and universities. The IRB protection of human subjects is of paramount importance given historical abuses (see Hamilton, 2005). Yet, the direct and indirect outcomes of the researcher IRB relationship have received limited scholarly attention and when they have, they have frequently been presented negatively as controversial or adversarial. Voiced Michael W. Kramer is at University of Missouri-Columbia. Vernon D. Miller is at Michigan State University. Suraj Commuri is at State University of New York*Albany. Michael Kramer can be contacted at [email protected] ISSN 0363-4523 (print)/ISSN 1479-5795 (online) # 2009 National Communication Association DOI: 10.1080/03634520903006208

498 M. Kramer et al.

concerns argue that the regulations are often too loose or too restrictive, the requirements are inconsistent from one university or department to the next, and the interactions with IRBs vary from professional to adversarial (Center for Advanced Study, 2005; Kramer & Dougherty, 2005). The nature of the researcherIRB relationship has a number of important consequences. For example, researchers’ trust in IRB directives and IRB’s confidence in researcher compliance provides the basis for the cooperative relationships needed to protect human subjects. The ease of understanding instructions and avenues to request clarification facilitate timely action. Perceptions of goodwill and intent shape attributions and how the IRB process is discussed with colleagues and students. As such, the nature of the researcherIRB relationship is of interest to communication scholars since the relationship is inherently a communicative one which relies (a) on both parties’ initiative in following guidelines and reporting their actions and (b) on transparency and ease of information exchange to prevent problems from developing. Moreover, the researcherIRB relationship fundamentally shapes the researchers’ attitudes and behaviors, which in turn influence the attitudes and behaviors of the researchers’ graduate and undergraduate students. Through social influence processes, faculty members, in their roles as supervisors and educators, socialize their students into the appropriate attitudes and beliefs to be participating members of the university culture (Jablin, 2001). Their messages help form colleagues’ and students’ perceptions of the IRB’s role in research (e.g., Narrative #34, 2005). Researchers who frame IRB compliance as a necessary evil or counterproductive to research goals convey a very different socialization message to students than researchers who portray IRB relationships as productive. To begin to examine these important issues empirically rather than anecdotally, this study systematically explored the relationship between IRBs’ communication and researchers’ attitudinal and behavioral responses from the researchers’ perspective. Review of Literature What follows is not a history of IRBs since that is available elsewhere (e.g., Hamilton, 2005). Instead, the following is a brief discussion of some of the important issues that have developed recently in the way IRBs operate at United States’ universities with a particular focus on those that potentially affect researchers’ attitudinal and behavioral responses to IRBs. These issues are perhaps unique to the U.S. since human subject research is governed differently in other countries.1 In addition, a variety of organizational communication literature is briefly reviewed since it suggests how communication between IRBs and researchers may influence attitudes and reactions. Institutional Review Boards Institutional Review Boards were created to protect human subjects from abuse by researchers in response to a number of (in)famous studies. Besides the atrocities conducted in the name of research by the Nazis during World War II, perhaps the

Communication and IRBs

499

most commonly referenced studies are Milgram’s (1974) psychological studies in which participants appeared to give excessive electric shocks to others without knowing that they were being deceived and the Tuskegee medical studies in which available medical treatment was withheld from individuals suffering from a sexually transmitted disease. Protecting human subjects from these types of abuses is an important and necessary function of IRBs. Despite the agreed upon positive purpose for IRBs, in recent years a variety of scholarship has portrayed IRBs negatively. For example, a number of scholars have questioned the general purview of IRBs suggesting that there has been a possible shift in focus from protecting human subjects to protecting universities from potential lengthy and expensive lawsuits by disgruntled human subjects (Annas, 2001; Foucar & Annas, 2002). For example, when legal counsel becomes involved in IRB adjudications of studies that eventually did not need informed consent, it suggested that a primary IRB goal may be protecting the university, not necessarily human subjects (see Dougherty & Kramer, 2005). Gunsalus et al. (2006) indicate that over 63% of citations by federal agencies for research violations are related to missing written procedures, poor minutekeeping, or failure to have a quorum at a meeting, rather than harming human subjects. For example, the University of Alabama-Birmingham was never charged with harming human subjects, just failure to follow rules, and the Office of Human Research Protection’s (OHRP) solution was to increase the bureaucracy at UAB by creating a new human subjects position (Brainard, 2004). Some scholarship casts IRBs in a negative light by suggesting they have become increasingly bureaucratic. IRBs initially started with small boards and a few guiding rules to follow. However, according to a recent Faculty Burden Survey, time spent complying with IRBs and their protocols was the burden most commonly listed by faculty (Kean, 2006). The increase in paperwork is particularly burdensome for federally funded research in which IRBs become involved prior even to the submission of grant proposals. As another indicator of increased bureaucracy, IRBs have been accused of increasing their oversight into additional areas. A recent white paper (Center for Advanced Study, 2005) explores the pros and cons of added regulations such as labeling research as risky when it asks human subjects to recall events in their lives, something which occurs in everyday conversations. Research protocols involving no interaction with humans, such as examining existing nonmedical databases and metaanalyses, are now receiving oversight from some IRBs. This steady increase in oversight, frequently referred to as mission creep, may reflect changes in safety concerns, but may also reflect zealous goals rather than protection of human subjects. Other scholarship suggests that one outcome of the researcherIRB relationship can be a sense of powerlessness among researchers. In response to a series of narratives about IRB experiences in a special issue of the Journal of Applied Communication Research (JACR) dedicated to IRB issues (Kramer & Dougherty, 2005), Fitch (2005), p. 276) wrote, ‘‘The sense of complete powerlessness, particularly from junior faculty, to question IRB decisions took me aback. I probably suffer from the usual blindness of people in power to how that looks and feels to the ones subjected to it.’’ The sense that the IRB decisions are irrevocable, possible to appeal but practically futile in terms of


weeks of delay and paperwork, make compliance the only viable option, especially when graduate students must complete a study during a semester or junior faculty members are working against a promotion and tenure clock. The importance of researcher attitudes toward IRBs as part of the educational process is highlighted by the fact that in the published JACR narratives, 76% of the contributors referred to discussions of IRB regulations between researchers and their students, and 14% disclosed that they have been noncompliant with IRB regulations. The decidedly negative portrayals of IRBs presented by the scholarship discussed so far are largely based on anecdotal evidence like the JACR narratives. For a more comprehensive and less biased perspective, it is important to examine the role of IRBs as a regulatory agency within its own larger organizational university setting, including the broader organizational communication dynamics. To that end, the next section explores the IRB using three different conceptual frames to consider the communicative, affective, and behavioral elements of the regulatory relationship. Bureaucracy, Communication Climate, and Quality of Communication Exchanges As a permanent structure at U.S. universities, IRBs fit classical definitions of bureaucracy (Weber, 1914/1978). They have fixed jurisdictions with codified rules and regulations which they are expected to enact in a stable and unbiased manner. Bureaucracies are a force of stability, consistency, and uniformity in organizations which are designed to create equitable treatment based on objective application of rules. Despite these potential advantages, research has found that such formalized structures often have a negative relationship to the frequency of various forms of communication (Jablin, 1987). Such a negative relationship is less likely to occur if the organization’s communication climate is described in terms of openness rather than closedness. Openness-closedness is a multidimensional construct of the communication climate that exists at the dyadic (between superiors and subordinates), group or unit (between leaders and group members), and organizational (between powerful elites and members) levels (Falcione, Sussman, & Herden, 1987; Miller, 2001). With regard to this study, IRBs demonstrate openness by acknowledging criticisms, responding in a nonhostile manner to suggestions, and inviting feedback. Conversely, IRBs convey closedness when they are perceived to dictate procedures or standards, provide inadequate feedback, and act in an arbitrary manner. Research in other organizations suggests that researchers who perceive an open communication climate with IRBs are likely to follow IRB rules and guidelines, express satisfaction with the IRB system, and generally support IRB policies (Falcione et al., 1987; Ruppel & Harrington, 2000). When researchers perceive a closed climate, they are likely to complain to peers about IRB practices, view IRB actions negatively, and report dissatisfaction with IRB policies and judgments. Researchers are also more likely to communicate those same negative attitudes to their graduate and undergraduate students as part of their education concerning conducting research. These affective responses likely influence behavioral responses including compliance for the current and next generation of researchers.


501

In addition, as regulatory agencies over researchers in their same organization, IRBs are in positions of power comparable to supervisors to subordinates. The research on supervisor-subordinate communication suggests a number of communication issues that likely shape researchers’ attitudes and behaviors toward IRBs. Subordinates respond positively to leaders who are more communication-minded, more willing to listen, ask or persuade rather than order, pass on information from their supervisors, and provide feedback (Jablin, 1979). Supervisorsubordinate relationships characterized by collaborative rather than directing communication, participatory rather than authoritarian decision making, as well as mutual trust and respect are associated with increased satisfaction and commitment (Graen, 2003; Graen & Uhl-Bien, 1995). Research Questions The research on bureaucracy, communication climate, and supervisor-subordinate communication suggest a number of communication elements that likely affect the faculty-IRB relationship. For example, if IRB policies and procedures are clear and directly relate to protecting human subjects, researchers are less likely to perceive IRBs as bureaucratic. If in their interactions with researchers, IRB staff members are consistent in their actions and open to feedback, researchers will perceive a more open than closed communication climate. If IRB staff members seem cooperative and respond in a timely manner, they will be perceived as more communication-minded in their regulatory role. Unfortunately, the JACR IRB narratives are replete with examples of IRBs that fall short in many of these areas. Together, documents like the Illinois White Paper (Center for Advanced Study, 2005) and the special issue of JACR (Kramer & Dougherty, 2005) suggest that IRBs’ communication and procedures are widely perceived negatively, that researchers are frequently dissatisfied with them, and that students are taught how to interact with IRBs by faculty. However, researchers’ perceptions and reported behaviors have not been systematically examined. In order to better understand that nature of the attitudes and beliefs that researchers hold and may be explicitly or implicitly teaching to their graduate and undergraduate students, the first question explored this: RQ1: Based on their communication interactions, what general attitudinal, affective, and behavioral reactions do researchers associate with IRBs at universities?

Since institutions and researchers are not homogenous, certain characteristics of the institutions or researchers may influence their perceptions and responses to IRBs. For example, researchers under pressure to publish in order to earn promotion and tenure at research institutions likely respond more negatively to delays and slow communication from an IRB than faculty members at small colleges where there are no such pressures. Since these differences may influence the attitudes and behaviors that researchers convey to their undergraduate and graduate students in different contexts, the second question explored these issues:


RQ2: Are researcher characteristics (e.g., age, type of school, research area, number of research protocols) associated with researchers’ perceptions of communication, satisfaction, or compliance with IRBs?

Beyond simply describing the perceptions of communication with their IRBs, it is important to explore the influence of these perceptions on researchers. This is an important issue because undergraduate and graduate students often emulate their mentors and form their likes or dislikes for IRBs and their procedures based on their faculty mentors. The research on communication climate and supervisor-subordinate communication suggests that certain types of communication will affect researchers’ satisfaction with IRBs (Falcione et al., 1987; Graen & Uhl-Bien, 1995). Thus, the third question explored affective responses to IRB interactions: RQ3: Which perceptions of communication behaviors by IRBs predict researchers’ satisfaction with their IRBs?

In addition to affective responses, Gunsalus et al. (2006) suggest that IRBs that are perceived as bureaucratic and unresponsive may actually cause researchers to become so alienated by the process that they become scofflaws; that is to say, they may respond like the author of Narrative #30 (2005) and actively circumvent IRBs, thereby potentially putting human subjects at risk. Since the behaviors demonstrated by researchers may be imitated by undergraduate and graduate students who are taught by them, the fourth question explored behavioral responses of researchers: RQ4: Which perceptions of communication behaviors by IRBs predict researchers’ compliance with the IRBs?

Method Respondents Respondents were recruited primarily through two methods. First, an e-mail list was created of authors of journal articles from four major communication journals representing the National Communication Association (NCA) and the International Communication Association (Communication Monographs, Journal of Applied Communication Research, Human Communication Research, and Journal of Communication). These journals were targeted because they publish research more likely to involve human subjects than journals that focus on other research methodologies. Each author with an accessible e-mail address received an invitation to participate in the study. In addition, general requests that researchers participate in the study were posted on CRTNET (electronic newsletter of NCA), a number of divisional listserves, and the organizational behavior listserve for the Academy of Management. Each e-mail or posting encouraged respondents to pass on the information to other researchers. A total of 426 respondents accessed the appropriate website and completed the survey online. The respondents represented a range of institutions including bachelor degrees only (9.6%), Masters degree granting (22.2%), and Ph.D. granting research institutions (66.5%), with a few indicating other types (1.7%). Respondents ranged in age from 21


503

to 90 (M 41.75; SD10.84) with 57% female and 43% male. The majority were communication scholars (66.4%), but other disciplines were also represented including business/management (15.3%), social/behavioral sciences (12.2%), and the rest not specifying (6.1%). They reported submitting an average of 4.34 (SD15.6; median3; mode2) IRB proposals per year and publishing from 0 to 300 referred articles in their careers (M21.3; SD33.34; median10; mode3 or 4). They reported that 85% of their research involved human subjects (SD23.88), 52% was conducted on campus (SD35.69), 46% involved students (SD34.47), 15% was grant-funded (SD24.71), and less than 1% of the protocols involved medical procedures (SD5.85). Survey Development Since measures for communication with IRBs, attitudes toward research, and outcomes were unavailable, new scales were developed for this study. Based on the IRB and organizational communication scholarship reviewed above and personal experience, items were created to represent the issues identified. These included perceptions of bureaucracy, mission creep, and lack of focus on the protecting human subjects, as well as specific problems like slow responses. These included 24 items to address various communication topics such as timeliness of responses, appropriateness and clarity of information requests, positive and open communication, consistency of regulations and procedures, and ease of application. In addition, 16 items addressed attitudes toward research such as need for protection, risk of research, and department attitudes. Finally, 12 items addressed three main outcomes, satisfaction, communication compliance, and behavioral compliance with the IRB. Items included both positively and negatively worded statements to reduce response bias and were presented in randomized order. Responses to these Likert-type scales ranged from agreeing ‘‘to a very little extent’’ 1 to ‘‘to a very great extent’’5. Items were examined for clarity and comprehensiveness by two former IRB members and then pretested on a small group of graduate students. After making adjustments based on the feedback, the survey was established on a website where anyone with World Wide Web access could complete it. To prevent participants from responding more than once, the survey program placed a ‘‘cookie’’ on the browser used to complete the survey that prevented future access to the survey.2 Measurements Three separate factor analyses (principal components, varimax rotation, eigenvalues 1) were conducted, one for the items describing interactions with the IRB, one for attitudes about research and the IRB, and one for the criterion variables. Items were retained if they loaded at greater than .50 on one factor and less than .40 on other factors. Reliability analyses (a) were conducted. Items which reduced scale reliability were dropped. Scale items and primary factor loadings are available from the first author. Interactions with the IRB. Factor analysis results indicated that 18 items represented three communication interaction factors. The first factor, importance, an eight-item


scale (a .92), measured the degree to which respondents considered the information requests for completing the IRB process important and useful rather than unproductive or frivolous (e.g., Completing the IRB applications is a good use of my time). The second factor, timeliness, a five-item scale (a.89), assessed respondents’ perceptions of the promptness of IRB reviews (e.g., The IRB communicates their decisions on my proposals in a timely manner). The third factor, consistency, a fiveitem scale (a.87), measured the degree to which respondents felt that policies and requests for information remained about the same over time (e.g., The IRB uses consistent interpretations of policies in making decisions on my research proposals). Attitudes toward research. Factor analysis for attitudes toward research indicated that 15 items represented three factors. The first, protection, a seven-item scale (a.82), indicated the degree to which respondents felt the IRB was necessary to protect human subjects from researchers (e.g., Human subjects involved in research on my campus would be at serious risk without IRB approval). A second factor, minimal risk, a four-item scale (a .80), measured whether respondents felt that their own research put human subjects at little or no risk (e.g., The human subjects in my research are at minimal risk). The third factor, department attitude, a 4-item scale (a.76), indicated respondents’ perceptions of how supportive and compliant their departmental colleagues were with IRB policies (e.g., Colleagues in my department carefully follow all IRB policies and procedures). Criterion variables. Factor analysis indicated that 10 items represented the three criterion variables. A four-item scale (a.92) measured general Satisfaction with the IRB (e.g., Overall, I am satisfied with my campus IRB). A three-item scale (a .75), Communication Compliance, measured the extent to which researchers reported providing all relevant information to the IRB (e.g., I provide the IRB with a complete explanation of my research proposals). Finally, a 3-item scale (a.92), Research Compliance, measured researchers’ practice of collecting data only following IRB approval (e.g., I never begin collecting data from people without IRB approval). Results Research Question 1: Reactions to IRBs Table 1 reports scale means and the correlation matrix for the study’s variables. A series of t tests determined that each scale’s mean differed significantly from the scale midpoint (3.0 or neutral on a 15 scale) (see Table 2). These results suggest that respondents’ attitudes were slightly positive about IRB policies being communicated in a consistent and timely manner, but slightly negative about whether IRB procedures requested information that was important. Their attitudes were slightly positive about the need for IRBs to protect human subjects and their departments’ support of the IRB. They felt that their own research was of minimal risk. They were slightly satisfied with their IRBs. They reported that they were generally compliant


505

Table 1 Scale Means, SD, and Correlations* Scale 1 2 3 4 5 6

Importance Timeliness Consistent Protection Minimal Risk Department Attitude 7 Satisfaction 8 Communication Compliance 9 Research Compliance

M; SD 2.90; 3.39; 3.19; 3.56; 4.46; 3.09;

1

0.95 0.95 .64 1.02 .73 0.74 .58 0.60 .24 0.78 .49

2

3

4

5

6

.69 .36 .13 .40

.37 .06 .41

.39 .19

.07

7

8

3.44; 1.01 3.50; 0.93

.74 .35

.81 .20

.81 .28

.43 .21

.11 .06

.47 .28

.26

4.00; 1.11

.27

.21

.17

.23

.05

.28

.26

.29

*r .10, p B.05; r .15, p B.01.

with IRB regulations. Overall, these results indicated that these researchers had slightly more positive attitudes toward interactions with the IRB than negative ones. They were slightly satisfied with IRBs and were generally compliant with IRB information requests and followed IRB procedures. Research Question 2: Differences by Researcher Characteristics A series of analyses explored whether researcher characteristics influenced perceptions or behaviors relating to IRBs. Correlations between continuous demographic measures (age, number of proposals per year, number of publications, percent grant funded, percent involving human subjects, percent conducted on campus, percent involving medical procedures, and percent involving student subjects) and the predictor and outcome variables were generally not significant. Due to sample size, a number of correlations were statistically significant with correlations ranging from .10 to .13, but were practically inconsequential accounting for less than 2% of the variance. The singular exception was the relationship between percentage of research involving medical protocols and risk (r .22, p B.01). Despite the small number involved in Table 2 Scale Means, Standard Deviations, and T-Test Differences from Scale Midpoint Scale 1 2 3 4 5 6 7 8 9

Importance Timeliness Consistent Protection Low Risk Department Attitude Satisfaction Communication Compliance Research Compliance

M; SD 2.90; 3.39; 3.19; 3.56; 4.46; 3.09; 3.44; 3.50; 4.00;

0.95 0.95 1.02 0.74 0.60 0.78 1.01 0.93 1.11

T-test difference from Scale Midpoint 2.02 8.36 3.75 15.31 49.42 2.22 8.83 11.08 18.49

pB.05 pB.001 pB.001 pB.001 pB.001 pB.05 pB.001 pB.001 pB.001


medical research, this correlation indicated that those engaging in more medical protocols were also more concerned about potential risk to human subjects in their own research. As a result, the percent of medical procedures was entered as a control variable in the following regression analyses. Results from analyses of categorical demographic variables revealed significant differences for type of institution (Ph.D. versus non-Ph.D.) on perceptions of departmental attitude, t(396) 3.56, h2 .03, pB.001, and research compliance, t(404)2.86, h2 .02, p B.01. Those in Ph.D. institutions (M 3.19) perceived their department as more compliant with IRBs than those in other institutions (M 2.89) and themselves (M 4.11) as more compliant with IRBs than those in other institutions (M3.78). As a result, type of institution was entered as a control variable in the following regression analysis. A series of ANOVAs found no significant differences based on researcher’s subject area. A series of t-tests found no significant differences due to researcher sex. Research Question 3: Predictors of Researchers’ Satisfaction with IRBs Based on RQ2 results, the percentage of medical procedures research and institutional type (Ph.D.non-Ph.D.) were entered as control variables in the first step in regression analyses with IRB satisfaction as the outcome variable. Then the remaining predictor variables were inputted under the stepwise procedure in step 2. The results reported in Table 3 indicate that the control variables entered in Step 1 were not significant predictors of researchers’ satisfaction with IRBs in the final analysis. However, variables entered in Step 2 significantly predicted satisfaction, accounting for 81% of the variance. In particular, researchers’ perception of timeliness (b .45), consistency (b.37) and importance of procedures (b .18) all positively predicted satisfaction. These results indicate that communication interactions with the IRB are directly associated with researcher satisfaction. Research Question 4: Predictors of Researchers’ Compliance with the IRBs Two separate regressions were conducted, one for communication compliance and one for research compliance using the same pattern as in the previous analysis. For communication compliance, the control variables entered in Step 1 were not significant predictors in the final analysis. The predictor variables entered in Step 2 were significant predictors of communication compliance, accounting for 14% of the variance. The more strongly researchers felt that IRBs requests for information were important (b.24), the more likely they were to report communicating openly with IRBs. Those who felt that their departments complied with IRBs policies and procedures (b.20) were also more likely to report communicating openly. For research compliance, the control variables entered in Step 1 were again not significant predictors in the final analysis. However, variables entered in Step 2 were significant predictors of research compliance, accounting for 13% of the variance. In particular, researchers who felt that their departments complied with the IRBs


507

Table 3 Stepwise Regression Results for Satisfaction and Compliance

Variables

Satisfaction b, p

Communication compliance b, p

Research compliance b, p

Step 1 Ph.D./Non-Ph.D. Percent medical

.01, p B.85 .01, p B.80

.07, p B.21 .05, p B.33

.09, pB.10 .07, pB.20

Step 2 Importance Timeliness Consistency Protection Low risk Department attitude F df r2 pB

.18, p B.001 .45, p B.001 .37, p B.001 291.3 5, 338 0.81 0.001

.24, p B.001 .20, p B.001 14.38 4, 342 .14 .001

.17, pB.001 .26, pB.001 15.02 4, 339 .13 .001

(b .26) and who felt more strongly that IRBs were needed to protect human subjects from researchers (b .17) were more likely to report complying completely with IRB research procedures. Discussion This study examined the attitudes and behaviors of researchers in relation to communication with IRBs. It found that on average researchers had neutral to slightly positive attitudes toward IRBs. They felt that their research involved little or no risk to human subjects whom they could protect without IRBs. Researchers reported being more satisfied with their IRBs if their communication occurred in a timely and consistent manner and included requests for information they felt was important in protecting human subjects. Researchers were more likely to report communicating openly and complying with IRBs if they felt their departments supported the IRBs and perceived the information requests as important. Researchers were also more likely to report complying if they believed that IRBs were necessary to protect human subjects from researchers. Overall, participants in this study seemed slightly satisfied with their interaction with IRBs. In general, they were compliant with IRBs by communicating openly and conforming to the research requirements. An important finding in this study was that the general attitudes toward IRBs appear to be fairly neutral to slightly positive. This contrasts the predominately negative narratives in the JACR special issue (Kramer & Dougherty, 2005) in which 36 of 57 (63%) narratives expressed negative experiences with IRBs and the negative tone of the Illinois White Paper (Center for Advanced Study, 2005). This likely indicates that instead of IRBs consistently creating serious problems for researchers,


the negative and positive experiences seem to balance out over time across IRBs and universities. This finding suggests that either IRBs at specific universities create significant problems or that researchers have negative experiences occasionally with IRBs that temper what are generally uneventful interactions. Although some issues need to be addressed across IRBs, these results suggest that the focus should be on solving problems with specific IRB staff members, with IRBs at particular universities, or with certain types of research protocols. Of concern, however, is the indication of a significant amount of noncompliance with IRBs. Although at first glance, noncompliance does not appear to be a problem since the means for the compliance scales are fairly high, careful examination of the results reveals that noncompliance is potentially a significant problem. For example, although the mean for research compliance was 4 (SD 1.11), an examination of the distribution indicates that approximately 24% of the respondents marked neutral (3) or lower on the scale. With the lower mean (M3.5; SD.93), an even larger percentage, 35%, marked neutral or lower for communication compliance. These results seem consistent with the 14% of the writers of the published JACR narratives who openly admitted to being noncompliant. There is no way of knowing the level of risk these researchers pose for human subjects. No doubt many do not pose a threat because their research protocols involve minimal if any risk to human subjects, but it is not clear how many may be unknowingly putting human subjects at risk. For example, the author of Narrative #28 (2005) did not appear to recognize the risk created by the breach of confidentiality created by a mass e-mail to participants. Other noncompliant researchers may be creating similar or more serious risks. Changes in communication with IRBs that can increase compliance will reduce these risks to human subjects. Of concern for the educational process are the attitudes and behaviors that researchers are communicating to their undergraduate and graduates students. This is a critical issue given the frequency with which researchers interact with students about IRBs (e.g., 76% of the published JACR narratives). Fortunately, despite concerns that some individuals are modeling noncompliant behaviors, results from this study suggest that most researchers are likely modeling and teaching their students to be compliant in their communication with IRBs and to follow IRB regulations. This seems quite consistent with the JACR special issue in which researchers felt powerless to do anything other than comply with IRBs and enacted very little resistance despite expressing negative attitudes and experiences (Dougherty & Kramer, 2005). That said, it should be a concern that most researchers’ attitudes toward the IRBs are rather neutral instead of positive and supportive. Improvements in the IRB process and communication would likely result in more positive researcher attitudes, which would also increase the probability of faculty members communicating more positive attitudes to students. Taken together, these results suggest two sets of communicative relationships between researchers, IRB members and others in the academic community. These relationships are presented in Figure 1 and could be used to guide future research. The first set of relationships identifies the influence of faculty-IRB interactions, researcher attitudes toward protecting human subjects, and researchers’ orientations

Communication and IRBs Messages Exchanged with Colleagues

509

Research Attempted

Researcher Attitudes

Faculty-IRB Interactions

Self-Protecting DegreeofRisk DepartmentAttitudes

Timeliness Consistency Importance

Researcher Orientation Researcher Satisfaction Communication Compliance Research Compliance

Messages Exchanged with Graduate and Undergraduate Students

Relationship to Other Regulatory Agencies

Figure 1. Model of facultyIRB relationships and university communication climates.

toward the IRB. These relationships in turn shape the second set of relationships, the nature of researchers’ message exchanges with colleagues, students, and other regulatory bodies, and the nature of research which they attempt. For example, researchers approach IRBs with certain attitudes about the need for protecting human subjects, the risk of their research, and their own department’s support of the IRB. Institutional Review Boards in turn communicate with a certain amount of timeliness, importance, and consistency. The interaction of researchers’ attitudes and IRB communication lead to the researchers’ orientation toward the IRB including satisfaction and compliance. These interactions and attitudes then influence a number of important additional behaviors of researchers such as the communication involved in the socialization of undergraduate and graduate students, messages to colleagues, research attempted, and most likely interactions with other regulatory agencies within the university and perhaps beyond. These behaviors, which may be positive or negative, then affect future attitudes and communication interactions with the IRB. This model should be useful in further investigation of a range of researcher IRB relationships and outcomes. Although the nearly neutral means and their standard deviations on the scales suggest some problems, the results of this study as represented in the model also suggest ways to improve researchers’ attitudes and compliance. More timely communication from IRBs to researchers suggests bureaucratization is not a problem and predicts more positive attitudes toward IRBs. When requests for information on IRB protocols seem important for protecting human subjects, it suggests the focus is on the primary mission of the IRB rather than on increasing bureaucracy. When there is inconsistency


over time, it likely represents mission creep as new regulations are added and greater scrutiny is given to new research compared to similar studies in the past. Maintaining consistent rules and procedures likely creates a sense that the procedures are important for protecting human subjects, and predicts positive researcher reactions and compliant behaviors. Changing them for reasons that appear unrelated to protecting human subjects likely contributes to a sense that the procedures are not important to follow and may lead to researchers downplaying IRB compliance in graduate student training. The JACR narratives are replete with stories of delays, inconsistency, and adversarial interactions. Addressing these concerns appears particularly important since those who are less satisfied with their IRBs are also less compliant (r .26) and potentially communicate such attitudes to their students. Since these issues were associated with researcher satisfaction and compliance, addressing them is important for improving compliance particularly given the foundational role these researchers have in the education and socialization of undergraduate and graduate students on how to interact with IRBs. It may be that universities need to address the issue of timely reviews by providing more staff to manage the workload. However, slow reviews may also result from overregulation (bureaucratization) and inconsistencies (mission creep) which slow the review process as much as understaffing. Narrative #4 (2005) lamented copying consent documents from a previously approved study only to have the IRB request changes in them which seemed to have little to do with improving the protection of human subjects. Researchers could submit protocols that would be more quickly approved if the guidelines were more consistent across time and staff members so that less researcher and staff time was spent modifying protocols to meet inconsistent guidelines. Such consistency will also facilitate teaching and mentoring of students because good and bad models of IRB applications can be archived in departments for use in teaching courses and assisting researchers. Dougherty and Kramer (2005) discuss the amount of effort, including consulting legal counsel, which went into approving a study that eventually was listed as not needing consent. This type of time-consuming, unnecessary regulation of exempt studies slows down the review process for other protocols, as well, and brews negative generalizations about the bureaucratic processes which eventually hinder the efforts of faculty mentors to educate their students about successful practices for communicating with IRBs. Given the strength of the findings (R2 .82) related to satisfaction, it is imperative that IRB staff and board members consider the impact of their communication on researchers particularly since satisfaction is related to compliance (r .26). Currently IRB staff and board members are trained in OHRP rules and regulations for conducting research. Perhaps it is time for similar training of IRB staff and board members on how to communicate more effectively with researchers. Fitch (2005) recommends improving researcher and IRB interactions by focusing on communication. Improvements in communication would likely be associated with increased ratings of satisfaction and compliance by researchers which they would then model


511

for their students. New communication training for IRB staff and researchers may be as necessary as the current training in compliance with rules and regulations. Most researchers in this study believe that their research puts human subjects at minimal risk, that they would protect their human subjects without IRB oversight, and that IRBs are not that necessary for protecting human subjects from researchers. Although it is important to recognize that researchers may be unable to appropriately make such an assessment, it does point to a problem mentioned in the JACR special issue (Kramer & Dougherty, 2005) and the Illinois White Paper (Center for Advanced Study, 2005): It is problematic that IRBs apply the same monolithic guidelines and protective procedures that may be necessary for high risk research, such as medical research, to studies with minimal risks in other disciplines. Adopting more appropriate procedures for minimal risk research would likely improve approval times, increase a sense that the less burdensome paperwork that is required actually is important, and likely improve compliance with IRB regulations. Perhaps most importantly, it will eliminate the possibility that faculty mentors will suggest to their students that they are better equipped to protect human subjects than IRBs. While a faculty member with years of research experience may be capable of genuinely protecting human subjects, a student who seeks to model such an IRB-independent approach to human subject protection, such as the graduate student in Narrative #28 (2005), may be lulled into a false sense of adequacy, with a potential for unfortunate consequences to human subjects. Limitations The high percentage of communication scholars and the use of an online survey may limit the results’ generalizability. However, the lack of significant differences across disciplines, types of institutions, and types of research suggests that the findings may apply more broadly. Indeed, the participants were experienced researchers with a median of three IRBs submissions a year and a median of 10 publications to their credit. Since only 12 respondents in this study (2.8%) indicated using medical protocols, additional analyses on these individuals were limited. A subsequent analysis excluding those using medical protocols and no longer controlling for percentage of medical research revealed little change in the findings. Another analysis that included the 12 respondents, but did not control for percentage of medical research, also found results consistent with those reported earlier. However, since the percentage of medical protocols was significantly associated with perceived risk, future investigations should examine those who use medical procedures in order to explore the nuances of their relationships with IRBs, especially since greater scrutiny is associated with potentially invasive protocols. There may be other contextual factors about the research that could also be examined, such as research related to homeland security or other sensitive topics. A number of the behaviors in the model were not examined in the current study. The JACR narratives provide evidence supporting such relationships. In the narratives, researchers describe teaching students how to interact with IRBs (e.g., Narrative #34,


2005), discussing their interactions with IRBs with other researchers (e.g., Narrative #12, 2005), and avoiding certain research topics due to the difficulty of completing the IRB process (e.g., Narrative #3, 2005). Research designs that include advisor-student pairs or colleague groups are needed to explore these relationships more closely in future research. The study also did not explore the experiences and perceptions of IRB board or staff members, an area ripe for future research. Likewise, distinguishing between exempt, expedited, or full-board reviews may provide important insights that focusing on researchers’ general experiences failed to notice. It is interesting that researchers did not perceive departmental members as particularly supportive of IRBs (M 3.09), but were more likely to comply if they perceived the department as compliant. This important finding underlines the critical role that socialization and norms about IRBs play as senior faculty members help shape the attitudes and practices of junior faculty and graduate students. However, given the one-time sample and self-report data, it is impossible to know whether researchers perceived their departments as compliant and as a result complied, or whether researchers were compliant and simply projected their attitudes on colleagues. Further research can explore the importance of departmental and campus-wide attitudes and behaviors toward IRBs particularly since these attitudes likely influence the attitudes and behaviors that other faculty members and students develop. It may also be useful to track specific IRB submissions over time or discover researchers’ orientations toward IRB compliance following specific positive or negative experiences. Future Research In addition to the proposed model, a variety of theoretical lenses may provide additional insight into IRB interactions. For example, some of the results seem consistent with various expectancy theories. Expectancy violation theory indicates that violations of nonverbal expectations result in negative affect and responses (Burgoon & Hale, 1988; Burgoon & Burgoon, 2001). Negative emotions result when there is a discrepancy between the expected and actual rate of progress toward goals (Carver & Scheier, 1990; Mandler, 1975). Unmet expectations results in dissatisfaction and turnover (Wanous, Poland, Premack, & Davis, 1992). Slow responses (42%) and adversarial or rude interactions (19%), behaviors reported in many of the JACR narratives (Koerner, 2005), and the importance of timeliness in the results reported here suggest IRBs likely violate researchers’ expectations on various occasions in their communication interactions. Alternative theoretical lenses may be of value in exploring how the IRB structure can produce helplessness, compliance, and noncompliance among different individuals. For example, the theories of reasoned action (Ajzen & Fishbein, 1980) or planned behavior (Ajzen, 1991) might provide insight into how perceived norms (perceptions of departmental members’ views) and perceived behavioral control (or powerlessness) impact intentions to comply or ignore IRB regulations. Uncertainty (e.g., Kramer,


513

2004) and information (e.g., Afifi & Weiner, 2004) management theories may provide insight into the choices researchers and IRBs make as they deal with problems such as inconsistencies, delays, and perceived lack of cooperation. Sense making (Weick, 2001) may provide insight into how individuals understand the process and their compliant or noncompliant behaviors. Finally, structuration theory (Giddens, 1979) might be useful for examining how researchers and board members both participate in the production and reproduction of the bureaucratic system of rules and resources of the IRBs. In addition to possible theoretical extensions, this research suggests that it would be valuable to explore communication in other regulatory relationships. Within and between organizations there are relationships that are similar to IRB relationships with researchers. Within organizations, departments must comply with policies and procedures implemented by human resources or accounting departments. Restaurants and contractors must comply with the rules and regulations enforced by government officials. Communication viewed as consistent, important, and timely may similarly affect satisfaction and compliance in these relationships. The results here suggest that developing an open- rather than closed-communication climate (Ruppel & Harrington, 2000) between regulators and those needing their approval will likely improve outcomes. The more the relationship can be described as similar to open, supportive supervisor relationships (Graen, 2003), the more likely individuals will cooperate and comply with regulators in protecting citizens. This study provides some possible starting points for exploring such regulatory relationships in other organizational settings. Conclusion This study provides one of the first systematic examinations of the communication between researchers and IRBs. As such, it provides an initial understanding of some of the important issues that affect the attitudes and behaviors of researchers as they interact with IRBs. It provides a snapshot of some of the ways in which concerns over these interactions can be addressed through improved communication. Given the importance of conducting research as part of faculty members’ roles at an increasing number of universities, and their influence in teaching research methods classes and in directing other research activities as educators and mentors in the socialization of undergraduate and graduate students who become future researchers, this study, along with additional research, can help make the process of generating new knowledge in that role more manageable and smooth for professors. At the same time, it can help create a safer environment for the human subjects who participate in that research. Notes [1]

A number of international scholars indicated that they were unable to participate, since research in their country was not governed by anything like IRBs.


[2]

Placing a ‘‘cookie’’ on the browser did not guarantee that individuals could not repeat the survey. Computer-savvy respondents could delete the ‘‘cookie’’ or use a different browser or computer to participate more than once. There is no evidence in the data that suggests individuals completed multiple surveys. We expect that as practising researchers, respondents’ own ethical standards would have restrained them without ‘‘cookies.’’

References Afifi, W. A., & Weiner, J. L. (2004). Toward a theory of motivated information management. Communication Theory, 14, 167190. Annas, G. J. (2001). Reforming informed consent to genetic research. JAMA, 286, 23262328. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50, 179211. Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicating social behavior. Englewood Cliffs, NJ: Prentice Hall. Brainard, J. (2004, April 13). Federal regulators again push U. of Alabama-Birmingham to improve human subjects protections [Electronic version]. The Chronicle of Higher Eductiaon, Today’s News. Retrieved April 13, 2004, from http://chronicle.com/daily/2004/04/2004041201n.htm Burgoon, J. K., & Burgoon, M. (2001). Expectancy theories. In P. Robinson & H. Giles (Eds.), Handbook of language and social psychology (2nd ed., pp. 79101). Sussex, England: Wiley. Burgoon, J. K., & Hale, J. L. (1988). Nonverbal expectancy violations: Model elaboration and application to immediacy behaviors. Communication Monographs, 55, 5879. Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positive and negative affect: A control-process view. Psychological Review, 97, 1935. Center for Advanced Study. (2005, November). The Illinois white paper: Improving the system for protecting human subjects: Counteracting IRB ‘‘mission creep.’’ Champaign: University of Illinois. Dougherty, D. S., & Kramer, M. W. (2005). A rationale for scholarly examination of institutional review boards: A case study. Journal of Applied Communication Research, 33, 183188. Falcione, R. L., Sussman, L., & Herden, R. P. (1987). Communication climate in organizations. In F. M. Jablin, L. L. Putnam, K. H. Roberts, & L. W. Porter (Eds.), Handbook of organizational communication: An interdisciplinary perspective (pp. 195227). Newbury Park, CA: Sage. Fitch, K. L. (2005). Difficult interactions between IRBs and investigators: Applications and solutions. Journal of Applied Communication Research, 33, 269276. Foucar, E., & Annas, G. J. (2002). How much oversight is necessary to protect human subjects? JAMA, 287, 716717. Giddens, A. (1979). Central problems in social theory. Berkeley: University of California Press. Graen, G. B. (2003). Role making onto the starting work team using LMX leadership: Diversity as an asset. In G. B. Graen (Ed.), Dealing with diversity: LMX leadership: the series, vol. 1 (pp. 128). Greenwich, CT: Information Age. Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership*development of leadermember exchange (LMX) theory of leadership over 25 years*applying a multilevel multidomain perspective. Leadership Quarterly, 6, 219247. Gunsalus, C. K., Bruner, E. M., Burbules, N. C., Dash, L., Finkin, M., Goldberg, J. P., et al. (2006). Mission creep in the IRB world. Science, 312, 1441. Hamilton, A. (2005). The development and operation of IRBs: Medical regulations and social science. Journal of Applied Communication Research, 33, 189293. Jablin, F. M. (1979). Superiorsubordinate communication: The state of the art. Psychological Bulletin, 86, 12011222.


515

Jablin, F. M. (1987). Formal organization structure. In F. M. Jablin, L. L. Putnam, K. H. Roberts, & L. W. Porter (Eds.), Handbook of organizational communication: An interdisciplinary perspective (pp. 389419). Newbury Park, CA: Sage. Jablin, F. M. (2001). Organizational entry, assimilation, and disengagement/exit. In F. M. Jablin & L. L. Putnam (Eds.), The new handbook of organizational communication: Advances in theory, research, and methods (pp. 732818). Thousand Oaks, CA: Sage. Kean, S. (2006, July 7). Scientists spend nearly half their time on administrative tasks, survey finds. Chronicle of Higher Education, A:23. Koerner, A. F. (2005). Communication scholars’ communication and relationship with their IRBs. Journal of Applied Communication Research, 33, 231241. Kramer, M. W. (2004). Managing uncertainty in organizational communication. Mahwah, NJ: Erlbaum. Kramer, M. W., & Dougherty, D. S. (Eds.). (2005) Communication research and institutional review boards [Special Issue]. Journal of Applied Communication Research, 33(3). Mandler, G. (1975). Mind and body: Psychology of emotion and stress. New York: Norton. Milgram, S. (1974). Obedience to authority: An experimental view. New York: Harper Perennial. Miller, K. I. (2001). Quantitative research methods. In F. M. Jablin & L. L. Putnam (Eds.), The new handbook of organizational communication: Advances in theory, research and methods (pp. 137160). Thousand Oaks, CA: Sage. Narrative #3. (2005). Journal of Applied Communication Research, 33, 204205. Narrative #4. (2005). Journal of Applied Communication Research, 33, 205206. Narrative #12. (2005). Journal of Applied Communication Research, 33, 213214. Narrative #28. (2005). Journal of Applied Communication Research, 33, 217219. Narrative #30. (2005). Journal of Applied Communication Research, 33, 221. Narrative #34. (2005). Journal of Applied Communication Research, 33, 221223. Ruppel, C. P., & Harrington, S. J. (2000). The relationship of communication, ethical work climate, and trust to commitment and innovation. Journal of Business Ethics, 25, 313328. Wanous, J. P., Poland, T. D., Premack, S. L., & Davis, K. S. (1992). The effects of met expectations on newcomer attitudes and behaviors: A review and meta-analysis. Journal of Applied Psychology, 77, 288297. Weber, M. (1914). Economy and society. Berkeley: University of California Press. Weick, K. E. (2001). Making sense of the organization. Malden, MA: Blackwell.