Skip navigation
College Quarterly
Winter 2012 - Volume 15 Number 1
Revisiting the AAUP recommendation: The viability of collegiality as a fourth criterion for university faculty evaluation
Pattie C. Johnston, Ph.D., Tammy Schimmel, Ph.D., Hunter O'Hara, Ed.D.
Abstract

Legal rulings have called for the inclusion of collegiality as a fourth evaluation category for university faculty. Collegialityis considered to be any extra-role behavior that represents individuals' behavior that is discretionary, not recognized by the formal reward system and that, in the aggregate, promotes the effective functioning of the educational organization. This article examines the important legal precedent that calls for the inclusion of collegiality as an evaluation category in light of a position taken by the American Association of University Professors (AAUP). The AAUP recommended against including collegial behaviors in faculty evaluations stating that the inclusion could hinder academic freedom by not allowing for dissent and that the construct of collegiality is amorphous and prevents the creation of an effective tool available to evaluate collegial behavior. The additional case made by the non-collegial faculty member is that collegiality is not part of their formal contract and they are, therefore, not held accountable for any lack of it. Faculty members have been denied tenure based upon a lack of ability to work well with others and have taken university decisions to courts. Courts have suggested that universities include collegial behaviors as part of the formal contract. Inclusion could put departments on better legal grounds for dismissal and may act as a deterrent for faculty denied tenure or promotion to pursue legal action. Studies have also shown that collegial behaviors are correlated with institutional effectiveness and valued by tenure track faculty. This paper will examine the legal findings may help guide university faculty when making tenure and promotion decisions. Additionally, the amorphous nature of the construct of collegiality will be delineated for more effective use in communicating expectations of collegiality.

Introduction

Teaching, service and scholarship are the three areas university faculty members are typically contracted to provide evidence of in tenure and promotion reviews. Faculty members also have additional extra role responsibilities that are not included in the formal contract. The extra-role behaviors referenced may be considered to be elements of collegiality. Specifically, collegiality may refer to any extra-role behavior that represents individuals' behavior that is discretionary, not recognized by the formal reward system and that, in the aggregate, promotes the effective functioning of the educational organization (Organ, 1988). The terms collegial behaviors and extra role behaviors are used interchangeably for the purpose of this paper.

There have been compelling assertions from relevant literature and legal findings that support university faculty members being held accountable for non-contractually bound collegial behaviors in tenure and promotion evaluations (Connell & Savage, 2001). Faculty members are also beginning to acknowledge the importance of collegial behaviors in position statements and other documents (Boyce, Oates, Lund & Fiorentino, 2008). The case for including extra role behaviors is further supported by research that investigated the perceived faculty importance of collegiality over other job factors such as salary and workload in the workplace (Fogg, 2006).

Other research suggests that there are problems associated with the inclusion of collegiality in faculty evaluations.  One problem with including collegiality is that the professoriate is not in agreement with having this fourth criterion. The American Association of University Professors (AAUP) has provided an influential voice on this issue by adopting a recommendation that does not support including collegiality as an evaluation criterion (AAUP, 2006). Perhaps the more overarching problem associated with the inclusion of collegiality is that critics have noted the amorphous nature of the construct which makes the development of an efficient and useful tool for its evaluation difficult (Connell & Savage, 2001). It would be difficult to hold faculty accountable for collegiality if specific performance expectations related to collegiality were informal and ambiguous.

Response to AAUP Recommendation

The AAUP's Committee A on Academic Freedom and Tenure has recommended that the practice of adding a fourth criterion of collegiality is discouraged (AAUP, 2006). The committee acknowledges the need for collegiality in the triumvirate of teaching, scholarship and service but noted several reasons why it should not be a separate entity. Several points mentioned in the AAUP recommendation described below warrant careful consideration.

The AAUP recommendation does call for "professional misconduct or malfeasance" which refers to non-collegial behaviors, to be relevant matter for faculty evaluation. They also note that efforts to obstruct the ability of colleagues to carry out their normal functions, engage in personal attacks or to violate ethical standards are actions faculty need to be accountable for in their evaluations but that they do not constitute a discrete standard. It is difficult to understand fully what is meant in the AAUP recommendation that states breaches in collegiality should be relevant matter. It suggests accountability for collegial behavior but it is difficult to hold faculty accountable in a meaningful way for anything that is not part of a formal contract.

The accountability for a collegial breech may currently be found in a contract other than the formal contract. Rousseau (1995) suggests that there is another type of contract that exists between colleagues which serves to bind extra-role collegial behaviors. She uses the term psychological contract to describe the subtle presence of expectations regarding terms of an exchange agreement between individuals and their organization. Arguyis (1990) defined the psychological or implied contract as the unwritten expectations that operate between people in a work place. Such a contract might include extra role collegial behaviors. The power of the psychological contract may be recognized when it is violated (Rousseau, 1989). The violation may be seen as more than just a failure to meet expectations but rather a signal of a damaged relationship between individuals. Bies (1987) suggests that perceived violation may yield deep and instant responses similar to anger and moral outrage. Current evaluation practices appear to be using implied contracts to "cover" mutual agreements and obligations about collegial behaviors assumed in typical written contractual agreements. The implied contract has been described as being a powerful determinant of behavior. However, court opinion has said the assumption of collegial expectations in the implied contract may leave universities legally vulnerable if dismissals are based on them (Schien, 1980). Examples of legal findings are described below.

Legal considerations. The AAUP recommendation says in its rationale that collegiality should never be the sole cause of non-reappointment, denial of tenure or dismissal and therefore, not be considered a discrete category. The difficulty that arises with this perspective is that the lack of collegial behavior has been and continues to be a sole criterion for dismissals and denials in numerous court cases.  Fisher (Fisher v. Vassar College, 1995) was denied tenure because she had difficulty in establishing straightforward, open, trusting, collegial relationships with others in the department. She brought suit against the college, was awarded damages but the federal appellate court overturned the decision and sustained Vassar's decision to deny Fisher tenure. The assenting opinion said that senior members of the department simply did not "like" Fisher and did not wish to establish a career-long professional relationship with her even though it was arguable that such grounds alone justified the department's recommendation against tenure.

Yackshaw was a tenured professor at John Carroll University who was found to have written an anonymous letter charging several members of his department with sexual harassment, mental illness, improper sexual conduct and homosexuality (Yackshaw v. John Carroll University, 1993).  He denied the writing of the letter but a history of similar behavior suggested otherwise and he was dismissed. The letter was said to have represented "moral turpitude". Yackshaw brought suit against the university, won, but lost in appellate court. The case of Kelly v. Kansas State Community College (1987) resulted in the dismissal of two nurses on the basis of their refusal to cooperate with other colleagues. Other faculty referred to the "constant snipping" in the staff meetings. The two nurses were dismissed, filed suit, but lost.

A professor at Hobart and William Smith Colleges was denied tenure citing problems in all three typically included criteria in the faculty handbook (Romer v. Hobart & William Smith Colleges, 1994).  The dean had written a letter to the review committee expressing her concerns about Romer's personal and professional relationship with another department member suggesting a collegial breach.  Romer filed a grievance which was rejected and then filed suit alleging breach of contract. It was his contention that the college considered this relationship when making their decision and that this type of behavior was not within any of the three stated criteria. The federal court rejected the notion that listing specific criteria in the faculty handbook somehow limited the types of information that a college could assess in its tenure process.

The case of Bresnick v. Manhattanville College (1994) involved another tenure denial because of lack of collegiality. Bresnick was denied because of his "inability to work with others." The college further stated that New York laws set forth for awarding tenure encompassed collegiality or working with colleagues in a collaborative manner. Bresnick's argument was that collegiality was not part of the criteria listed in college documents and therefore, he could not be dismissed on these grounds. The court again deferred to the college but stated that extra role responsibilities should be part of the contract.

In each of these cases, court decisions have deferred to university judgment but the rulings continue to recommend formal inclusion of expectations. Despite the AAUP assertion that collegial issues should not be the sole basis for denial of tenure or dismissal, these cases confirm that they have been. More problematic for universities is that professors who are denied tenure or promotion may be more likely to continue to bring suit against departments and universities as they have in the past if they think that grounds are present because collegial expectations are not articulated.

Perceived threat to academic freedom. Perhaps the strongest argument the AAUP and its supporters have made for not including collegiality as a separate criterion of faculty assessment is that the invocation of collegiality may threaten academic freedom. The AAUP (2006) stated that collegiality may be confused with the expectation that a faculty member display "enthusiasm." "dedication," "evince a constructive attitude that will foster harmony" or display "excessive deference to decisions that may require reason." Such expectations are flatly contrary to principles of academic freedom which protect a faculty member's right to dissent from judgments of colleagues and administrators. This concern is serious but may be reconciled when indicators associated with or representative of collegiality are validated. The delineation of valid collegiality indicators could enable universities to accept collegial behaviors expected by faculty and reject any behaviors that would threaten academic freedoms.

Supporting Organizational Behavior Literature

The importance of collegiality in the university workplace has been underscored in the literature. Bodies of research from the educational evaluation and industrial organizational literature suggest that collegiality be addressed because of its link to organizational effectiveness.  Trower and Gallager (2008) suggest that it is vital for institutions to value and implement accountability for collegial interactions. They say the valuing of collegiality is especially important for chairs to embrace so that they can become effective mentors and relationship builders as their position may be used as an impetus for creating collegial workplaces.  The evidence provided by these two bodies of literature is both empirical and descriptive in nature which strengthens the argument for inclusion.

Descriptive support. Collegiality is closely related to other constructs named in educational and organizational research to include rapport, pro-social organizational behaviors and organizational citizenship behaviors. Tickle-Degnen & Rosenthal (1987) said that rapport was important to the coordination of behavior in social interactions but emphasized that it did not require a cheerful attitude. It is not considered a character trait but persons may be particularly skilled at developing it. Pro-Social Organizational Behavior is a model of organizational extra role behaviors that represents the types of non-contractually bound behaviors expected to occur between co-workers (Brief & Motowildo, 1986). Staw (1986) expanded the behavioral model and delineated 13 indicators of pro-social organizational behaviors. Sample behaviors include assisting co-workers, suggesting improvements, objecting to improper directives, putting forth extra effort and volunteering for extra assignments. Organ (1988) has provided a model of Organizational Citizenship Behaviors (OCB) that has received extensive attention in the organizational literature. The model has five separate categories of extra role behaviors which include Altruism, Conscientiousness, Sportsmanship, Courtesy and Civic Virtue. Altruism includes behaviors that have the effect of helping a person with an organizationally relevant task such as lending materials. Conscientiousness includes behaviors that go beyond a minimum requirement such as when existing faculty members help to orient a new professor. Avoiding negative behaviors such as petty complaining encompass the third category of Sportsmanship. Courtesy includes briefing colleagues and other behaviors performed to prevent problems. Civic Virtue requires the responsible participation in the political life of the university or organization. Civic Virtue is closely related to service but encompasses more than behaviors typically associated with service. Behaviors may include reading relevant materials or discussing issues. The attention that extra-role behaviors have received in the research serves to underscore their importance in the workplace and is further supported by empirical studies.

Empirical evidence. There is also considerable empirical evidence that has suggested the importance of collegiality in the workplace. The organizational studies have used the term "pro-social behaviors" to describe extra role or collegial behaviors. Studies have investigated the relationship between extra role behaviors and overall organizational effectiveness. Connell and Savage (2001) have asserted that a person's ability to work civilly in the university is no different from acting civilly outside the academy which suggests that findings from organizational research have application to university settings in terms of organizational effectiveness. George, Bettenhausen, & Rausen (1990) investigated pro-social behavior and group performance and found that these two variables were significantly related (r = .33). They also looked at the relationship between group cohesiveness and pro-social behaviors. Highly cohesive groups were characterized by heightened collegial behaviors including member attraction to the group, friendliness, mutual liking and positive feelings about carrying out group tasks. Correlations calculated indicate a high linkage between cohesiveness and pro-social behaviors which suggests that extra-role behaviors not only impact productivity but relate to other desirable aspects of organizations such as cohesiveness (r = .50). Puffer (1987) also looked at the relationship between pro-social behaviors and work performance and found an equally positive relationship between the two variables.

Podsakoff and MacKenzie (1994) examined the relationship between these two variables. They specifically used OCB categories as their metric for pro-social behaviors. Correlations between performance effectiveness and each OCB category were calculated. Results indicated that all categories were found to be related to overall performance which supports findings mentioned above that suggest extra role behaviors are important to organizational functioning. Civic Virtue and Sportsmanship exhibited expected positive effects on performance. Helping behavior was found to have a negative effect on performance possibly because the effect of helping may have been statistically suppressed. Podsakoff, Ahearne and MacKenzie (1997) also conducted a study that examined the effectiveness of OCBs and addressed the possible statistical suppression from the previously mentioned study. Data provided support for the hypothesis that all OCBs with the exception of Civic Virtue were related to effectiveness including helping behaviors. The authors suggested that perhaps the effects of Civic Virtue showed up only in aggregate which makes accountability difficult.

These findings suggest that collegiality related behaviors benefit organizational effectiveness in substantive and positive ways. The descriptive support is rooted in significant educational evaluation writings as well as in the above mentioned writings in organizational behavior. The empirical evidence has suggested a relationship between extra-role behaviors and organizational effectiveness. Both branches of research provided compelling reasons to consider faculty accountability.

Faculty Viewpoints of Collegiality

The significance of collegiality to university faculty is also supported by current educational research. A national survey created by the research center at Harvard University's Graduate School of Education was administered to 4,500 tenure-track faculty members from 51 colleges and universities. Survey data found that tenure-track faculty members care more about departmental climate, culture, and collegiality than they do about workload, tenure clarity, and compensation (Fogg, 2006). A regression analysis of the responses indicated that climate was five times as important as compensation in predicting job satisfaction. These results have significant implications for retaining tenure track faculty. If departments want to retain effective tenure track professors, then this study suggests that they could provide proactive support for collegiality skills.

In addition to faculty perceived importance of a collegial atmosphere, Boyce, Oates, Lund and Florentino (2008) have noted that trends have already revealed the emergence of collegiality as a fourth category of formal assessment in some institutions. They state that "dispositions and collegiality" are appearing in promotion and tenure guidelines and position announcements as characteristics of successful candidates. They also note that efforts to assess faculty affective behaviors have not resulted in the development of an effective and useful tool. Performance expectations have been informal and unwritten which has created an evaluation environment that is ambiguous.

Other researchers have claimed that collegiality is such an amorphous term that it could be subtly and adversely used in evaluation of minority faculty (Connell & Savage, 2001). The suggestion is that faculty could use a lack of collegiality as a mask for discrimination or faculty distaste for other faculty with opposing viewpoints. Even though critics of collegiality have expressed concerns about the possible discriminatory misuse of collegiality, the courts have decided in favor of the universities in almost every case where the issue was raised.

The amorphous and ambiguous nature of the term collegiality is also relevant to the concern raised by Boyce, Oates, Lund and Florentino (2008) which noted that there was no effective tool for assessing extra role behaviors. An effective tool for evaluating collegiality would necessitate reducing the amorphous nature of collegiality by identifying valid indicators of collegiality. Evidence of indicator validity could be provided by using traditional measurement techniques for delineating constructs including review by subject matter experts (Crocker & Algina, 1986).

Purpose

The support for the addition of collegiality in faculty evaluations is compelling and therefore, warrants further development. However, there were two factors to consider before implementing a system that evaluates extra role behaviors. It has been noted that the term collegiality is so amorphous and vague that it does little to provide specific guidelines for behaviors and yet specific descriptors of collegiality have been identified and described in the above-mentioned studies. The AAUP has also warned that any assessment of collegiality should not inhibit academic freedoms. These two hindrances are reconciled by the careful delineation of collegiality as a construct.  One purpose of this study was to delineate collegiality by creating a list of possible research based indicators and assessing evidence of their validity. The list of validated indicators could be used to develop an effective tool for assessing collegiality as suggested by the courts and make legal actions taken in response to the lack of collegiality easier. The indicators would help create a clearer and less amorphous lens of collegiality and provide specific guidelines to help better communicate extra-role expectations between faculty members. Validated indicators could also serve to address the academic freedom concern of the AAUP. Faculty examination of each possible indicator is crucial to this effort so that indicators that may hinder academic freedoms are identified and rejected. Indicators are for use in Research I and II Universities only.

Method

The process of delineating the construct of collegiality required two basic procedures: the creation of a collegiality model with behavioral indicators and the implementation of a Job Task Analysis (JTA) to assess the validity of the indicators by having a national random sample of professors rate the representativeness of the suggested indicators.

Creation of Initial Collegial Model

The study required the construction of a model of collegial behaviors representative of university professors from Research I and II universities. The preliminary nature of this attempt required that all behavioral indicators derived from the research on collegiality be included for professor rating and examination in order to enhance validity by capturing the domain more completely. Indicators were derived from the OCB model, the list of prosocial organization behaviors, the aspects included in the model of rapport and all individually listed components from the education evaluation research previously cited. Each indicator was subsumed by one of the five categories of OCB to enhance organizational salience. These categories were adapted to the academic setting. These possible indicators of collegiality were to be reviewed by university professors and rated as to their representativeness of expected collegial behaviors.

Research Participants

This study involved the participation of two groups of professors. The first group of research participants consisted of four Research I University faculty members from three different departments functioning as Subject Matter Experts (SMEs). They were asked to review the initial list of possible collegial behaviors before the list was sent out to a large sample of professors. The SME's were asked to judge the representativeness of all possible collegiality indicators. They were also asked to augment the list with other possible indicators of collegial behavior not included. It was of particular concern to have one of the SME's from the psychology department who specialized in organizational behavior. Another SME functioned as a department chairperson and added an administrative perspective. The two remaining SME's were from various departments based on availability. Tenured and non-tenured professors were included.

The second group of research participants consisted of a nationwide random sample of Research I and II University professors. A table of random numbers was used to select the sample. The sample was selected by first breaking down the universities into major fields of study. The major fields of studies used were adapted from the program fields used by The National Center for Educational Statistics in a survey conducted by the U.S. Department of Education (1993).All major fields of study were sampled in this project and include Art, Business, Social Science, Communication, Computer Science, Foreign Language, Education, English, Engineering, Natural Sciences, Mathematics, Psychology, Philosophy, Religion and Theology. Each field of study was common to all Research I and Research II universities. Some fields were omitted because they were not common to the research universities. Law, Vocational Training and Agriculture were examples of some of the omitted fields.

A random sample was drawn to include professors from four Research I and four Research II universities for each of the 13 fields of study. The approximate sample size was estimated to be just under five hundred. Norgen and Krejcie (1970) have provided a table of recommended sample sizes for populations with finite sizes. Two hundred seventeen professor responses were needed according to these authors to generate an appropriate sample size. The survey was sent to the sample of professors who were asked to rate the degree to which each indicator represented collegial behavior.

Instrumentation

There was one initial survey of collegial behaviors that went through revisions as the study progressed. Identified indicators were listed under the appropriate OCB section for organizational saliency. Each of the initial 47 indicators was listed with a place to rate the degree to which the item is representative of collegial behavior. A Likert type format was used with five response options ranging from very indicative to not indicative of collegial interaction. The survey was given first to the abovementioned group of SMEs to examine. The SMEs rated the degree of collegiality of each indicator as well as augmented the list with additional indicators. Revisions were made to reflect the SME input and then sent out to the nationwide sample of professors. A cover letter was included with the survey to explain the purpose of the survey and to encourage participation. A follow-up letter was also sent to increase the response rate.

Another form was also created and given to the same SMEs to review each of the final indicators for racial/gender bias and ambiguity. The bias and ambiguity form was a Likert scale that included all indicators. SMEs were asked to rate the degree to which bias and ambiguity were present.

Implementation of Job Task Analysis

A Job Task Analysis (JTA) was conducted to provide evidence of the validity of each indicator by asking a large, national sample of university professors to rate the representativeness of each indicator. Clifford (1994) has suggested this methodology when trying to investigate relevant aspects of varying jobs by asking job incumbents to rate job aspects using a Likert scale. The JTA was modified for use in this study asking incumbents to investigate representativeness of collegial job behaviors. A JTA allowed for each possible indicator of collegiality to be rated by a large sample of individual professors without influence. Individual and uninfluenced ratings are important to the identification of representative indicators especially when considering status aspects. It is especially essential that minority and non-tenured faculty review collegiality indicators without pressure from tenured colleagues because critics of collegiality have expressed concern that these groups may be adversely affected by collegial evaluation. Therefore, their uninfluenced ratings of collegiality are warranted. Related Delphi Techniques purposely seek consensus which was not the desired outcome in this study and do not easily allow for the inclusion of large and random samples.

Clifford (1994) has provided a generic model approach to JTA which includes four steps. The first step includes collecting relevant job task data. This research effort used the extra-role behaviors delineated in the model as job tasks. The list of possible extra-role tasks was augmented by behaviors indicated as important from the Subject Matter Experts (SMEs). These experts provided information regarding the validity of the indicators (Crocker & Algina, 1986). They were asked to examine the list of extra-to-role behaviors and judge their representativeness of collegiality using a five point Likert scale. Five response options ranging from very indicative to not indicative at all were included. The data resulting from this judgment task were used for selection of items for the subsequent version to be used. Any items receiving a rating of "not indicative at all" by all four experts were dropped. All other items remained. The SMEs were also asked to augment the existing list with any indicators they felt were relevant.

The next step required the rated representativeness of the items to be verified by exposure to a large number of job incumbents. This step provided an opportunity for a large nationwide sample of professors to rate the appropriateness of the indicators. Kirwan and Ainsworth (1992) have suggested several ways to collect data. Activity sampling, critical incident technique, observation, questionnaires, interviews and verbal protocols were all possible employable techniques. Questionnaires or surveys are recommended if responses from a large number of incumbents are desired which was appropriate for this study. The questionnaire consisted of the list of revised collegial behavioral tasks based on the input from the four SMEs. These surveys were sent out to a large national random sample of professors from Research I and II Universities. The final step of the JTA involved analyzing the data received from the incumbent's questionnaires. Mean scores were calculated for each indicator. The means for appropriateness suggested which indicators appear to represent collegial behaviors best. Indicators with mean scores under 3.0 were dropped. Effect sizes between gender and tenure status were also calculated. Following Cohen (1988), effect sizes larger than .50 were considered to be too large and those items were removed. Items were also removed if suggested by several respondents.

Results

A model of collegiality was created based on research from relevant educational and industrial organizational literature that included 47 possible indicators subsumed by one of the five OCB categories. A group of four SMEs were given a copy of the survey for review before it was sent out to the national sample of professors. The SMEs comments/suggestions and ratings were reviewed. Revisions made in the wording of the items were based on suggestions made by the SMEs. The words "as needed" were added to assists co-workers with problems. The word "positive "was added to the two items about "contact with departmental co-workers." "When helpful" was added to briefs co-workers and the phrase "important to functioning" was added to attends meetings regularly. Of notable relation to the hindering of academic freedoms, the item "objects to improper objectives" was changed to include "politely objects." The item states that behaviors associated with academic freedom such as objecting or dissenting is acceptable behavior but that it should be done politely (please see Table 2 for complete items).

Mean ratings and range of ratings were calculated for each of the 47 indicators for revision review. Item 27 was the only item dropped because of a low mean score (2.75). Ninety eight percent of the items were rated above the 3.0 level which indicated a rating of somewhere above moderately indicative of collegiality. Score ranges were small which suggested that SMEs were in approximate agreement in their ratings.

The revised survey was then sent to a nationwide random sample of over 730 Research University I and II professors. The sample size exceeded the estimate needed. Three hundred and one surveys were returned after two mailings which provided a 41% overall response rate. Analysis was conducted to determine which items were rated high enough to be representative of collegiality. Item means were calculated from both mailings and compared. Results suggested that mailing did not impact findings. Items with low means, large gender, race or tenure status effect sizes or at least two instances of repeated negative faculty input were dropped. Items were dropped if the mean approximated 3.0 because that suggested that the item was not indicative enough of collegiality.

Effect sizes were calculated to examine possible bias on items between race, male/female and tenured/non-tenured respondents. Cohen (1988) suggests that an effect size of .2 is small, .5 is medium and .8 is large. Any items with effect sizes nearing .5 were dropped. Three items were dropped because of large gender effect sizes (.46, .48 and .50). Each of these items was rated more representative of collegiality by men than by women. Two items were rated higher by tenure track faculty than tenured faculty (.55 and .47). No items were dropped based on effect sizes by race. Six items were reworded to enhance comprehension. Twenty two items were dropped and two items were added based on repeated incumbent suggestions.

Item means were calculated to determine which indicators were rated to represent collegiality well. Means approximating 3.0 were considered for dropping because this rating indicated that the item was "indicative of collegiality" on the Likert scale. Five items were dropped because their associated mean rating was at or below 3.0. These items included "does not leave organization despite hardships," "is not alienated from colleagues," "does not appear to be socially withdrawn," "does not lack social ability" and "speaks up" with respective means of3.05, 3.02, 2.81, 3.15 and 3.12.

Several items were combined because they were considered by at least two responders to be too similar to be separate indicators of collegiality. "Avoids petty grievances" and "does not make large problems out of small ones" were combined. "Touches base" and "provides advanced notice" were also considered to be too similar and combined into one item, "touches base."  Two items that were focused on faculty dissent were combined and included "exhibits good tension" and takes "turns in conversation." The suggestion was also given to make the phrase "good tension" less ambiguous and so the item was changed to "negotiates respectfully with co-workers." The last items that were combined are "attends meetings regularly" and "participates in casual conversation relating to work."

The following items were dropped because at least two incumbents provided similar kinds of suggestions. "Provides informal counseling to students" was dropped because it was about student/faculty interaction and not colleague interaction. "Represents organization favorably to others" was also dropped because it was considered to be an ethics issue. "Is compatible with others" was dropped because it was thought to be too general.

Six items were reworded to lessen ambiguity. "When approached or appropriate" was added to "assists co-workers with personal problems." "Willing to teach undesirable courses" was changed to "share of undesirable courses." The term "generally" was added to "displays a positive attitude" and to "is not disruptive." The last item was reworded to include "to department or college" when suggesting improvements. This change would exclude any personal suggested improvements.

Two items were added based on repeated incumbent written feedback requested at the end of the survey. All raters were provided space to describe any indicators that were not included in the survey but considered to be collegial behaviors. The first added item was based on six raters' comments on avoiding negative gossip about colleagues. The other item mentioned by three raters addressed praising the achievements of others.

Cronbach alphas were also calculated on each of the five OCB subscales to determine the internal consistency of the items (see Table 1). The alpha levels were high with the lowest level associated with the Altruism scale. The Altruism alpha level (.70) may be lower because of the small number of items on the scale. The items on this subscale had very high means which did indicate that the JTA sample found the indicators to be very representative of collegial behaviors.

The revised list of collegial behaviors included 27 items. The same group of SMEs were then given the list of indicators and an evaluation form and asked to evaluate the indicators for ambiguity, racial bias and gender bias. Most items remained the same. The wording was changed in two items. The phrase "when approached" was dropped from assisting co-workers with work related problems and the phrase "appropriate share" was added to volunteering for share of extra jobs. No items were rated as having any bias or as being unclear or ambiguous. The final list of indicators remained with 27 items (see Table 2). The Altruism, Conscientiousness, Sportsmanship, Courtesy and Civic Virtue subscales now had 4, 9, 5, 4 and 5 indicators each, respectively.

Discussion

Considerable evidence suggests that a fourth category, collegiality, be added to faculty evaluations despite recommendations by the AAUP and concerns about the amorphous nature of the construct.

A possible solution to the AAUP concerns about inhibiting academic freedoms and the vague nature of the construct was offered in this study that systematically created and validated a new model of the construct of collegiality. Professors rated an initial list of possible collegial indicators to determine the degree to which each indicator represented collegiality. The indicators could provide a source from which departments or colleges are able to move collegial behaviors into a more formal and effective assessment system by providing validated indicators. The indicators allow for clear delineation of faculty expectations which may be added to the evaluation system in a variety of ways.

Limitations

The indicators were rated by professors in Research I and II universities only. This limits the generalizability of the indicators to research universities because varying university environments may be associated with different collegial expectations. The limitations may be particularly problematic for universities and colleges with smaller department and faculty sizes or with less publication pressures. It could be that these types of colleges function with different degrees of interaction than large research universities.

The results are generalizable to all Research I and II universities which suggests that any national university of this type could apply these indicators. The larger than statistically needed sample size and randomness of the sample selection enable one to be fairly certain that the indicators demonstrated enough utility to call for further testing.

Implications

Approximately half of the original indicators were retained after the national sample of incumbent ratings. Most of the remaining indicators were revised to reflect collegial behaviors more carefully. All four of the retained indicators from the Altruism scale are about helping, assisting and consulting others. Feedback from professors indicated that these related behaviors were collegial but only "when needed" or "when asked." This distinction may be important because the implication is that helping and assisting colleagues without being invited is not collegial. A point not investigated in this study is whether who is helping whom matters. It may be that unsolicited help by a senior faculty member is different than unsolicited aide from a colleague of equal status.

Another item of note is the inclusion of the "negotiating well with others" indicator. This item reconciles concerns the AAUP has stated about academic freedom. The item implies that dissent is expected but that it should be done respectfully. Effort was made through several revisions to assure that the item was worded clearly and not open to interpretation.

Two indicators of collegiality not found in the literature were included in the final list of indicators. The addition of the indicator that refers to gossiping was of particular interest because it was strongly suggested without prompt in the open ended section of the survey by many incumbents. Gossiping negatively about peers was reported to be the cause of many faculty problems. The second item not suggested in the literature but suggested without prompt was praising the accomplishments of others. These items were added to the final list of indicators because of the repeated strong suggestions. Additionally, these two indicators appear to have face validity or have appearance of collegiality.

The indicators have practical use for departments. They could be part of a faculty handbook for reference or become part of a formal contract so that they may suggest better guidance. They could also provide documentation and positioning for the university if there is ever a legal dispute over extra-role behaviors. Each department may include varying behavioral expectations and could consider which of the indicators apply to their own department. The decision to augment or omit behaviors can be a collaborative and dynamic departmental decision based on consensus of collegial expectations. The list of collegial expectations may also be posted as part of a job description or discussed when interviewing any prospective candidate applying for a faculty position. Expectations can be made clear from the start.

Indicators of collegiality could also provide valuable and precise feedback for tenure track faculty. At present, junior faculty may be given only generalized references of collegiality providing junior faculty no precise guidance. Precise indicators clarify expectations for their behavior and contributions. The indicators may also help faculty leaders negotiate with senior faculty who see themselves as collegial but are not perceived by their peers as collegial. The indicators provide a clear expectation of collegiality by providing useful and exacting guidelines. Speaking of collegiality in vague terms may actually encourage some faculty to pursue what they believe to be collegial behaviors when, in fact, specific indicators will demonstrate otherwise.

One of the more important implications for faculty of this study may be a call to faculty leaders. Awareness by provosts, deans and chairpersons of the importance of extra role behaviors to the effective running of departments may be crucial for the valuing of a collegial atmosphere. These persons can be an important impetus for creating positive atmospheres and collegial work environments.

Future Research

There are several aspects of collegiality inclusion that could be examined with further investigation. The indicators could be turned into a scale to use for measurement of collegiality which could formalize the inclusion. Such a scale could be used by departments to provide formal feedback for non-tenured and tenured professors about exact behaviors in question. Scale construction would require several considerations. The final list of items/indicators was 27. This number is rather large. It may be beneficial to explore the factor structure in order to try to reduce the number of indicators and factors. There are several factors that may be reduced. Two sets of indicators of particular interest in this regard are: (a) assists colleagues with work related problems/consults with others on work related matters and(b) encourages faculty support/supports faculty. The implementation of the scale would require a departmentally salient plan to include who rates who, who is rated and when to rate. A scale may be of particular use in departments with a history of collegial problems. The two added items referring to gossip and giving praise for accomplishments may also warrant further validation.

Conclusion

The inclusion of collegiality as a fourth criterion for faculty evaluation has had two major problems. The AAUP issued an official recommendation against the inclusion of collegiality as a separate aspect of faculty evaluation. The recommendation was made primarily because holding faculty accountable for collegial behavior was thought to be a threat to academic freedom. The AAUP specifically said that including collegiality may be confused with an expectation that a faculty member display some sort of false enthusiasm, dedication or evince a constructive attitude just to foster harmony. This possibility is contrary to principles of academic freedom which say that a faculty member has a right to dissent from judgments of colleagues and administrators. The construct of collegiality also had not been delineated and associated indicators validated for university use. The amorphous nature of the construct made creating an effective assessment tool difficult.

These concerns are serious but were answered in the creation of a model of collegiality with validated indicators. Indicators of collegiality that reflect a need for respectful negotiation are included in the model of collegiality created in this study which alleviates the AAUP's concern. The use of carefully delineated indicators of collegiality would certainly help prevent the threat posed to academic freedom by rejecting any indicators that suggest the very types of behaviors cautioned by the AAUP in their recommendation. The amorphous nature of collegiality has been defined which will allow for the creation of an effective tool for its assessment and inclusion as a fourth criterion in faculty evaluation.

References

AAUP. (2006). Policy and Document Reports (10th Ed.), American Association of University Professors: Washington, DC.

Adams, M. (1989). Tenuring and promoting junior faculty. Thought & Action,5, 55-60.

Arden, E. (1989). Who should judge faculty? College Review Board, 152, 37-39.

Bateman, T. & Organ, D. (1983). Job satisfaction and the good solider: The relationship between affect and employee citizenship. Academy of Management Journal, 26, 587-595.

Bies, R. (1987). The predicament of injustice: The management of moral outrage. In Cummings, L. & Staw, B. (Eds.), Research in Organizational Behavior(35-72). Greenwich, CT: JAI Press.

Boyce, A., Oates, R., Lund, J. & Florentino, L. (2008). Faculty collegiality and dispositions in the tenure and promotion process: Developing a performance rubric. Academic Leader, 8. Bresnick v. Manhattanville College, 864 F.Supp. 327 (S.D.N.Y. 1994).

Brief, A. & Motowildo, C. (1984). Pro-social organizational behavior. Academy of Management Review, 10,710-725.

Clifford, J.P. (1994) Job analysis: Why do it, and how it should be done? Public Personnel Management, 23.

Cohen, J. (1988).Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum Associates.

Connell M. A. & Savage, F. G. (2001). Does collegiality count? Academe, 87, 37-40.

Crocker, L. & Algina, J. (1986). Introduction to classical and modern test theory. TX: Holt, Rinehart and Winston.

Fisher v. Vassar College, 66 F.3d.379 (2nd Cir. 1995).

Fogg, P. (2006). Young Ph.D.'s say collegiality matters more than salary. The Chronicle of Higher Education, 53, 1-4.

George, J., Bettenhausen, K. & Rausen, C. (1990). Understanding prosocial behavior, sales performance and turnover: A group level analysis in a service context. Journal of Applied Psychology, 7,698-709.

Kelly v. Kanas State Community College, 648 P.2d.1213 (Wash. 1987).

Kirwan, B. & Ainsworth, L. (1992). A Guide to Task Analysis. Bristol, PA: Taylor Francis.

Little, J. (1990). The persistence of privacy: Autonomy and initiative in teachers' professional relations. Teachers College Record, 91,509-527.

McKitrick, S. (2007, April). Assessing ‘ineffable' general education outcomes using the Delphi approach. Paper presented at the NCSU Assessment Symposium, Cary, NC.

Organ, D. (1988). The motivational basis of organizational citizenship behavior. In Staw, B. & Cummings,L. (Eds.) Research in Organizational Behavior(44-67). Greenwich, CT: JAI Press.

Podsakoff, P., Ahearne, M. & MacKenzie, S. (1994). Organizational citizenship behavior and the quantity and quality of work group performance. Journal of Applied Psychology, 82,262- 270.

Podsakoff, P. & MacKenzie, S.(1994).Organizational citizenship behaviors and sales unit effectiveness. Journal of Marketing Research, 31, 351-363.

Puffer, S. (1987). Pro-social behavior, non-compliant behavior and work performance among salespeople. Journal of Applied Psychology, 72, 615-621.

Romer v. Hobart & William Smith Colleges, 842 F. Supp. 703 (W.D.N.Y. 1994).

Rousseau, D. (1989). Psychological contracts in organizations: Understanding written and unwritten agreements. Thousand Oaks, CA: Sage Publications.

Schien, E. (1980). Organizational Psychology. Englewood Cliffs, NJ: Prentice Hall.

Staw, B.M. (1986). In pursuit of the happy/productive worker.  California Management Review, 28, 40-53.

Tickle-Degnen, L. & Rosenthal, R. (1987). Group processes and intergroup relations. In Hendrick, C. (Ed.), Review of Personality and Social Psychology. Newbury Park, CA: Sage.

Trowler, C. & Gallager, A. (2008). Why collegiality matters. The Chronicle of Higher Education, 3, 2-3.

Yackshaw v. John Carroll University, 624 N.E.2nd 225 (Ohio App. 1993).

Table 1: Subscale Cronbach alphas
Subscale  Cronbach alpha Number of items JTA respondent N
Altruism .70    4 274
Conscientiousness .80 9 283
Courtesy .93 5 291
Sportsmanship .92  4 279
Civic Virtue .86      5 285
Table 2: Final List of Incumbent Reviewed and Rated Indicators of Collegiality


Model of Collegiality by OCB Category

Altruism

  • Assists co-workers with job related problems
  • Assists co-workers with personal problems when needed
  • Shares materials when needed
  • Consults with others on work related problems when needed

Conscientiousness

  • Puts forth extra effort on the job
  • Serves on university wide committees
  • Volunteers for appropriate share of extra jobs or assignments
  • Agrees to teach an appropriate share of undesirable courses
  • Displays a generally positive attitude
  • Has positive contact with co-workers within own department
  • Has positive contact with co-workers outside of own department
  • Encourages faculty
  • Supports faculty

Sportsmanship

  • Avoids excessive complaining
  • Avoids petty grievances
  • Is not disruptive in meetings
  • Negotiates respectfully with co-workers
  • Praises achievements or awards of co-workers

Courtesy

  • Does not "gossip" negatively about co-workers
  • Challenges perceived injustices in a respectful manner
  • Demonstrates respect towards co-workers
  • Touches base with relevant persons

Civic Virtue

  • Regularly attends meetings important to departmental functioning
  • Promptly keeps appointments with co-workers
  • Completes committee responsibilities and assignments on time
  • Suggests improvements to the department or college
  • Contributes to joint efforts

Pattie C. Johnston, Ph.D., 401 West Kennedy Boulevard, Tampa, Florida. 33606 pjohnston@ut.edu