Dimensions of student perceptions of faculty ethical behavior: refining a measure and relationships with selected outcome variables.

Schnake, Mel ; Fredenberger, William ; Dumler, Michael P. 等

ABSTRACT

Scale items used in previous studies of student perceptions of faculty ethical behavior were administered to a sample of undergraduate business students in order to assess the dimensionality of these perceptions. Results suggest five distinct dimensions of student perceptions of faculty ethical behavior. These empirically derived dimensions were then related to three outcome variables, satisfaction with the university, intent to remain in the current degree program, and intent to transfer to another university. Results of this analysis suggest that one dimension of faculty ethical behavior is related to student satisfaction with the university, and that satisfaction, in turn, was related to student intent to transfer to a different university.

INTRODUCTION

Recent developments in business (e.g., Enron, Adelphia, Tyco, WorldCom, and Andersen Consulting) and politics (e.g., the expulsion of James Traficant from the House of Representatives, and the recent Letter of Admonition from the U.S. Senate Ethics Committee to Senator Robert G. Torricelli) has brought ethics to the forefront of public debate and media attention. However, while several scholars have suggested that teaching is rife with ethical dilemmas (Svinicki, 1994), relatively little empirical research has been conducted on ethics in academia (Tabachnick, Keith-Spiegel & Pope, 1991; Keith-Spiegel, Tabachnick, and Allen, 1993). The little research on faculty ethical behavior which has been conducted has typically focused on such topics as sexual harassment and has largely ignored the many daily ethical dilemmas which occur in faculty-student interactions (Tabachnick, Keith-Spiegel & Pope, 1991). Surprisingly, there is very little research on student perceptions of faculty ethical behavior, and the outcomes of those perceptions (Morgan & Korschgen, 2001). The few studies which have addressed student perceptions of faculty ethical behavior have ignored significant methodological issues, such as the dimensionality of the rating instrument used to assess student perceptions, and the validity and reliability of the measurement. The primary purpose of this study is to improve the measurement of student perceptions of faculty ethical behavior. Items from previous questionnaires related to a narrow range of faculty behaviors dealing with the professor-student interaction as well as additional items to more fully capture this domain will be examined for content validity, dimensionality, and reliability. Identifying dimensions of student perceived faculty ethical behavior will enable researchers to create a valid and reliable scale which can then be used to examine outcomes of student perceptions, such as student retention. Relationships between these empirically-derived dimensions of student perceived faculty ethical behavior and selected outcome variables (student course satisfaction and intent to remain in their current degree program or retention) will then be examined.

ETHICAL BEHAVIOR

Ethical behavior is rather difficult to define precisely given its origin in personal, organizational and societal values (Ball, 2001). Not that long ago, ethics was defined largely in terms of complying with the law. While, expectations have risen in recent years, it is difficult to draw a precise line between what is generally considered ethical behavior, and what is not. In a general sense, what is considered ethical behavior is what "most people" would consider right and wrong. Clearly, different groups may define ethical behavior differently. Therefore, one determinant of ethical behavior is the context within which it occurs. Ethical behavior in one industry or occupation may be quite different from another. Ultimately, an individual's ethics are his or her beliefs about what is right or wrong, good or bad (Garrett & Klonoski, 1992). Thus, an individual's actual ethical behavior is influenced by his or her personal beliefs and values, the climate of the organization, as well as society's values.

FACULTY ETHICAL BEHAVIOR

Some authors have expressed skepticism about the general ethical integrity of university and college professors (Anderson, 1992; Callahan, 1982; Sykes, 1989). Others point out how difficult it is to develop standardized expectations for behavior in university settings given the diversity, professional training, and independence of faculty (Whicker & Kronenfeld, 1994). A recent survey found that 74% of college seniors in a Zogby International poll reported that their professors, when teaching ethics, tell students that there is no clear right or wrong; rather, what is "ethical" depends on one's individual values and culture (Leo, 2002).

Yet other authors argue that ethical behavior has a key role in higher education (Smith, 1996; Little, 1989; Markie, 1994; Strike & Soltis, 1992). Evidence of the growing concern of ethics in academia is the number of professional organizations which have developed codes of ethics (e.g., the Academy of Management, American Association of University Professors, American Psychological Association). In addition, many colleges and universities have developed codes of conduct, although it has been suggested that many of these take a legalistic or minimalist approach, defining codes of conduct largely in terms of legal compliance (Kibler, 1994; Zabihollah, Elmore & Szendi, 2001).

There is some evidence that student perceptions of ethical climate may be related to important outcome variables such as student retention (Lipshutz, 1993; Parker, 1997). The 103-item Ethical Climate Index was administered to a sample of graduate students and faculty in the college of education in a Midwestern state university (Schulte, 2001). Schulte (2001) concluded that the data indicated that both faculty and students believed that a positive ethical climate is important in student retention in the degree program. Day-to-day faculty behavior and interactions with students would clearly influence student perceptions of the ethical climate, and thereby their likelihood of continuing in their current program of study.

MEASURES OF STUDENT PERCEPTIONS

Given the paucity of research addressing student perceptions of faculty ethical behavior, there are few research instruments available. Tabachnick, Keith-Spiegel, and Pope (1991) developed a 63-item measure which they used to compare the perceptions of teaching psychologists with clinical psychologists. The items were based on a previous study of teaching psychologists, and represent what the psychologists, not necessarily the students, consider to be examples of unethical behavior. The psychologists indicated the extent to which they engaged in these behaviors as well as rating how ethical they viewed each behavior. They also administered this instrument to college students and professors, asking them to rate the acceptability of behaviors in which professors might engage. They found great similarity between student and professor perceptions, male and female student perceptions, freshmen and upper-class student perceptions, and between Midwest and West Coast student perceptions.

In a follow-up study, Keith-Spiegel, Tabachnick and Allen (1993) developed a 107-item scale, taking 59 items from their earlier questionnaire (Tabachnick, Keith-Spiegel & Allen, 1991), and an additional 48 items from incidents contributed by students responding to a request to describe ethical problems with professors. Thus, approximately half of these items were identified by students as examples of faculty unethical behavior. Students rated these items in terms of how ethical they considered the behavior. They found few differences between freshmen and advanced students, between Midwest and West Coast students, and significant differences on only 6 of the 107 behaviors between male and female students. The types of faculty behaviors perceived as most unethical by students are those which impact fairness, and acts which violate a perceived class contract (e.g., "using a grading procedure that does not adequately measure what a student has learned," "giving every student an A regardless of the quality of their work," and "changing the criteria for successful completion of the class in the middle of the semester").

Birch, Elliott and Trankel (1999) developed a 64-item measure which asked respondents to rate the ethical appropriateness of 64 behaviors. Twenty-seven items in this scale were adapted from the Tabachnick et al. (1991) and Keith-Spiegel (1993) scale, and thirty-two items were developed from a review of other ethics surveys, from discussions with faculty and students, and from participant (faculty) discussions during a workshop on ethics. This instrument was administered to the faculty at a Western state university. They found that there was strong agreement among the respondents about fairness in grading, relationships between faculty and students, and university responsibilities. However, there were "gray areas" about which there was much less agreement concerning other aspects of student/faculty relationships. Similar to previous research, they found no differences between male and female respondents.

Morgan and Korschgen (2001) asked a sample of faculty and undergraduate students to rate the ethicalness of 16 faculty behaviors on a five-point scale with endpoints of "unquestionably not ethical" to "unquestionably ethical." They selected items from the Tabachnick, et al. (1991) scale which dealt primarily with student-faculty relationships. They found significant differences between faculty and student perceptions on four of the sixteen items. Faculty saw increasing popularity by giving easy tests, accepting a textbook rebate, sexual involvement with a student, and using profanity in lectures as more unethical than did students. Students saw using old lecture notes, and the breaking of students' confidences as more unethical than did faculty.

These instruments all measured a relatively broad array of faculty behaviors which were judged to have ethical implications. In addition to the day-to-day professor-student interactions (e.g., behavior in the classroom, assignments, grading), these questionnaires measured such faculty behaviors as dealings with colleagues, inappropriate use of the institution's resources, and avoiding college/university or departmental responsibilities.

One common methodological feature, and significant limitation, of these studies is the use of multiple item measures of perceived faculty ethical behavior, yet analyzing each item individually. All of these researchers conducted statistical tests of each individual item included in their scale. There was no assessment of discriminant or convergent validity or even reliability estimation. Nor was there any attempt to validate dimensions of perceived faculty ethical behavior across different groups of respondents (e.g., students and faculty). The problem with this, of course, is that some of these items may be highly correlated and therefore redundant. If many subjects interpret two items in a scale to mean the same thing, then we no longer have two separate, independent measures. We are not measuring two separate constructs, and they should not be interpreted nor analyzed as such. They should be combined into a single measure. This is particularly likely to occur when several similar items are presented to subjects at the same time. Respondents may honestly interpret several items similarly, or they may respond to several items in a manner similar to phenomenon known as common method variance. Respondents may, for example, attempt to be consistent with their responses, producing spurious, inflated correlations among the items. (Podsakoff & Organ, 1986). Regardless of the specific reason, the result is, in the respondents' minds, several items are the same. Analyzing or interpreting the results of these items individually is no longer appropriate.

A perhaps more significant problem with a single item measure is the difficulty in assessing its reliability and validity. In other words, it is difficult to establish precisely what is in fact being measured with single item measures.

Branstetter and Handelsman (2000) used a modified version of the Tabachnick et al. (1991) scale to measure graduate student perceptions of psychology faculty ethical behavior. In the first part of their study, they asked graduate students to evaluate how ethical they viewed 50 faculty behaviors. In the second part of the study, they asked respondents to report how frequently they engaged in the same 50 behaviors. They point out that the Tabachnick, et al. (1991) scale relates to six distinct areas of faculty behavior: (1) in-class issues, (2) lessons and evaluations, (3) outside-of-classroom issues, (4) relationships in academia, (5) responsibilities to students and colleagues, and (6) issues unique to teaching of psychology. This implies that the 63 items in the Tabachnick, et al. (1991) scale should collapse into these 6 dimensions. In fact, Branstetter and Handelsman (2000) did perform an exploratory factor analysis on the data from the first part of their study, which resulted in 5 factors (based on the eigenvalues = 1.00 rule of thumb and a scree test). Thus, the analysis did not support the a priori six dimensions of ethical faculty behavior. They interpreted these factors as follows: (1) systemic issues in academia; (2) dual relationship issues between graduate teaching assistants and students; (3) sexual relationships between faculty and students; (4) fidelity, justice and general competence in teaching; and (5) veracity, confidentiality and professional issues. A subsequent analysis examined graduate teaching assistant perceptions on these factors across age, gender, year in program, and specialty area. They found that older teaching assistants were less likely to rate dating and sexual relationships with students as ethical, and that, compared to women, men were more likely to rate the dual relationships factor as unethical but more likely to rate dating and sexual issues as ethical. Using data from the second part of their study, where they asked the same graduate teaching assistants to rate the frequency in which they themselves engaged in each of the 50 behaviors, the 50 items were compared individually across various subgroups (e.g., gender, age). One of their conclusions was that male graduate teaching assistants (GTAs) engaged in certain behaviors more frequently than did female GTAs. However, two of the three behaviors (scale items) loaded on the same factor (Factor 3: Sexual Relationships; item 9: "Becoming sexually involved with a student", and item 18: "Engaging in sexual fantasies about a student"). Since these loaded on the same factor in the principal components analysis, it makes little sense to further analyze them individually. In the respondents minds, these items measure largely the same construct. The third item included in this analysis did not load significantly on any of the factors in the exploratory factor analysis. Therefore, it is difficult to interpret what exactly is being measured with this item, or assess its validity and reliability.

SINGLE ITEM MEASURES

It has long been accepted that multiple-item measures are preferable to single-item measures (Nunnally, 1978). This preference is based on the general notion that a measurement instrument consists of a random sample of items from the domain of all items which measure the construct of interest. There are several other, more specific reasons for the preference for multiple-item measures. First, it is not possible to test the internal consistency reliability of a single-item measure. In addition, reliability can often be improved by adding items to a scale. However, recently it has been suggested that there are circumstances where single-item measures may be desirable ( Scarpello & Campbell, 1983; Wanous, Riechers, & Hudy, 1997; Gardner, Cummings, Dunham & Pierce, 1998; Nagy, 2002). The relative advantages and disadvantages of single-item versus multiple-item measures is not the focus of this research. Rather, this research examines the possible misuse of previous multiple item measures of student perceptions of faculty ethical behavior as single-item measures, and creating a valid and reliable scale which taps a limited part of the domain of faculty ethical behavior, specifically, those dealing with professor-student interactions. More completely specifying this domain by developing additional scale items and assessing the dimensionality, content validity and reliability of this instrument will enable researchers to more fully test the notion that student perceptions of faculty ethical behavior are related to important outcomes such as student retention.

RESEARCH GOAL

The purpose of this research is to examine a multiple-item measure of student perceived faculty ethical behavior to assess the dimensionality, or content validity, of this measure. The ultimate goal of this research is to develop a valid, yet more parsimonious, measure of this construct. This measure is based on items taken from the Tabachnick, Keith-Spiegel, and Pope (1991); Keith-Spiegel, Tabachnick and Allen (1993), and Morgan and Korschgen (2001) instruments. Items were selected which focused on the faculty-student relationship both within and outside of the classroom. There are clearly other important types of faculty behaviors which have ethical implications, however, students are not qualified judges of these behaviors (e.g., assisting junior colleagues). We focused on the limited domain of ethical faculty behaviors which students could readily observe and thus could accurately judge.

Empirically-derived dimensions of faculty ethical behavior will then be regressed on important outcome variables including student course satisfaction and student retention (intent to remain in their current degree program and intent to transfer to a different university).

MEASURES

A 95-item questionnaire was developed, taking items from previous measures (Tabachnick, Keith-Spiegel & Pope, 1991; Keith-Spiegel, Tabachnick & Allen, 1993; Morgan & Korschgen, 2001; Branstetter & Handelsman, 2000). Items were selected which dealt specifically with faculty behaviors relating to student-faculty interaction both within and outside of the classroom. Additional items were added to the scale to more completely measure the ethical principles related to student affairs suggested by Kitchener (1985), and the five areas of teaching behavior with distinct ethical implications suggested by Smith (1996). Kitchener (1985) suggested five ethical principles in dealing with students: (1) respect for autonomy, respect for the individual's right to make his/her own decisions; (2) nonmaleficence or doing no harm to others; (3) beneficence or benefiting others; (4) justice and treating others fairly; and (5) fidelity, being loyal and trustworthy. Smith (1996) suggested that certain faculty behaviors are directly related to ethical issues: (1) total student empowerment (e.g., turning over responsibility for the course to students, structuring the course to achieve high student ratings); (2) instructional materials and student expectations (e.g., choosing a textbook without reading it); (3) instructor's presence (e.g., frequently missing class, not being available to students); (4) examinations and grading (e.g., giving test questions not related to course material); (5) professional relationships (e.g., giving a high recommendation for a student whose performance does not warrant it). At this early stage in the research process, it was important to be inclusive in order to fully measure the domain of faculty ethical behaviors relating to student-faculty interaction and identify dimensions of this domain. This domain is illustrated in Figure 1.

[FIGURE 1 OMITTED]

Items included were developed based on both faculty and student inputs. The original Tabachnick et al. (1991) scale was generated based on input from teaching psychologists. However, the follow-up study (Keith-Spiegel, Tabachnick and Allen, 1993) included approximately the same number of items based on input from students. The Birch, Elliott, and Trankel (1999) scale was based in part on the Tabachnick et al. (1991) scale, but additional items (32 items) were generated based on discussions with both faculty and students. Thus, the scale employed in the present study is based approximately equally on input from faculty and students. In addition, there is recent evidence that students and faculty have relatively similar views regarding faculty ethical behavior (Morgan & Korschgen (2001).

Instructions on this scale defined ethical behavior in terms of right and wrong, and distinguished it from desirable and undesirable behavior (Appleby, 1990; Murray, 2000), as recommended by Morgan and Korschgen (2001). Respondents were instructed to focus on whether a behavior was right or wrong, and not whether they liked or disliked it. A seven-point Likert-type scale was employed with scale anchors of Definitely Ethical (1) and Definitely Unethical (7), with a midpoint of Neutral (4).

METHOD

The 95 item Faculty Behavior questionnaire was administered to several undergraduate business classes during the third week of class in the Fall Semester 2002 at a medium-sized state university in the south, and in a large lecture-hall management class at a large state university in the Midwest. A total of 132 students responded at the Midwestern university. The average age of these students is 21.03, 50.4% are male (.8% did not report their gender), and the average grade point average is 2.97. At the southern university,

The Faculty Behavior scale was submitted to a principal components analysis resulting in nineteen factors with eigenvalues greater than 1.0, explaining 65.85% of the variance in the scale. However, not all of these factors were interpretable, and a scree test suggested four to five factors. A varimax rotation was performed and items which did not load cleanly (factor loading >= .50) on a single factor were omitted. A second factor analysis was performed on the reduced-item scale (10 items were omitted), again producing nineteen factors with eigenvalues >= 1.0. A scree test again suggested between four and five factors. Once again, several factors were not interpretable, and several items did not load cleanly on a single factor. The data were submitted to a varimax rotation, and each scale item was examined to determine those items which exhibited very high (>= .50) loadings on a single factor. Thirty such items were identified.

These thirty items were submitted to a principal components analysis, resulting in five factors with eigenvalues >= 1.0, explaining 60.53 % of the variance. A scree test also suggested five factors. The results of this analysis appear in Table 1.

Nine items loaded cleanly on Factor 1. These items deal with a professor's behavior within the classroom, including grading, examinations, and fairness in dealing with students. This factor was interpreted as Classroom Ethical Behavior. Five items loaded on Factor 2. These items pertain to loaning money to students, borrowing money from students, accepting gifts from students, and asking favors of students. This factor was interpreted as Inappropriate Exchange. Six items loaded on Factor 3. These items deal with friendships and non-romantic relationships with students. This factor was interpreted as Relationships Outside the Classroom. Four items loaded on Factor 4. These items focus on dating and having sexual relationships with students. This factor was interpreted as Dating and Relationships. Six items loaded on Factor 5. These items have to do with effort and the use of time. More specifically, these items addressed the professor not being adequately prepared for class, not maintaining scheduled office hours, frequently arriving late for class, frequently missing class without notice, and taking a long time to grade and return exams. This factor was interpreted as Inappropriate Use of Time. Coefficient alpha reliability estimates ranged from .79 to .89 for these five factors. Reliabilities and correlations among the five empirically derived dimensions of faculty ethical behavior appear in Figure 4. Reliability estimates appear on the diagonal. As Table 2 shows, there is a moderate degree of correlation among the five factors. Factors 2 and 3 (favors and loans, platonic relationships outside of the classroom), factors 2 and 4 (favors and loans, dating), factors 3 and 4 (platonic relationships outside the classroom, dating), and factors 1 are 5 (classroom behaviors, inappropriate use of time) are most strongly related.

OUTCOME VARIABLES

Questionnaires containing outcome measures were administered to students in these introductory management and information systems classes. The outcome measures were administered approximately two months after the administration of the ethical behavior scale. Student ID numbers were used to match respondent questionnaires. A thirteen-item scale measured retention (intent to finish the current degree program at the current university) (9 items, e.g., "I plan to finish my current degree program at this university."), and satisfaction with the current university (4 items, e.g., "I really do not like this college/university"). Principal components analysis of these 13 items revealed three factors. Three retention items ("There is a very good chance that I will transfer to another college/university to finish my degree," "I plan to start looking at other colleges/universities to transfer to," and "I plan to change colleges/universities as soon as possible.") loaded on a single factor. This factor was interpreted as "intent to transfer." Three additional retention items ("I plan to finish my current degree program at this college/university," "I made the right choice when I enrolled in this college/university," "I'm very unhappy with my educational experience at this college/university.") loaded on a second factor. This factor was interpreted as "intent to remain." The four satisfaction items (e.g., "I am very satisfied with my educational experience at this college/university," "I really do not like this college/university.") loaded together along with the remaining three retention items.

Coefficient alpha reliability analysis was performed on the items making up these three factors. Reliability estimate of the intent to remain scale is .87. A coefficient alpha reliability estimate of .91 for the intent to transfer scale was achieved by dropping the "I plan to change colleges/universities as soon as possible" item. Finally, a coefficient alpha reliability coefficient of .81 was achieved by dropping the "I really do not like this college/university" item.

RELATIONSHIPS AMONG PERCEIVED FACULTY ETHICAL BEHAVIOR AND OUTCOME VARIABLES

Variables for the five faculty ethical behavior subscales and the three outcome variables were created by averaging respondents' scores on these items. The three outcome variables were regressed on the five faculty ethical behavior scales. The results indicate that only one of the faculty ethical behavior scales was significantly related to any of the outcome variables. Inappropriate Exchange was significantly and negatively related to satisfaction with the university (R2 = .06, p < .05; b = -.20, p < .01).

Two-way and three-way interactions among the outcome variables were tested and found to be non-significant. However, bivariate correlations of -.20 (p < .01) and .17 (p < .05) was observed between satisfaction and intent to transfer, and satisfaction and intent to remain, respectively. This raises the issue of whether student perceptions of faculty ethical behavior might have an indirect effect on retention through satisfaction with the university. To examine this question. additional regressions were performed. First, Inappropriate Exchange has already displayed a relationship with satisfaction with the university. Next, satisfaction was regression on intent to remain and intent to transfer. Satisfaction with the university was significantly related to intent to transfer (R2 = .05, p < .01; b = -.24, p < .01). Thus, while not providing a complete test of direct and indirect relationships, the results are suggestive that student perceptions of faculty ethical behavior may impact satisfaction with the university, which in turn affects students' intent to transfer.

DISCUSSION

Previous versions of measures of faculty ethical behavior have generally been rather lengthy (e.g., 63 items). However, this research provides evidence that several dimensions of student perceived faculty ethical behavior can be measured with a much shorter scale.

The only previous factor analysis of student perceptions of faculty ethical behavior (Branstetter & Handelsman, 2000) also identified 5 dimensions. However, this previous research employed a 50-item measure of specific teaching and teaching-related behaviors in a sample of graduate teaching assistants. There are some similarities between this previous research and the present study. Both studies identified a factor involving lending money to students, asking favors of students, selling goods to students, and accepting gifts from students. Both studies also identified a factor dealing with dating and romantic relationships between professors and students. In addition, both studies identified a factor related to fairness and general competence in teaching (e.g., using alcohol or drugs while teaching, not maintaining scheduled office hours, missing class).

This study has identified five distinct dimensions of student perceptions of faculty ethical behavior, and conducted initial work toward developing a more parsimonious measure of these dimensions. Further psychometric work on scale development is warranted to further validate this measure. Once the scale is finalized, it can easily be re-written to assess the "ethical climate" of the institution by asking respondents to rate the frequency of each behavior in general among the faculty. However, a reliable and valid measure of this construct should provide the impetus for examining how student perceptions of faculty ethical behavior, or the "ethical climate" of the institution, affect important outcome variables such as student retention, class attendance, perceived stress, and even student performance.

This study also provides some preliminary evidence that student perceptions of faculty ethical behavior may be related directly and/or indirectly to some important outcome variables, including satisfaction with the university, intent to remain, and intent to transfer. While faculty ethical behavior in the present study was significantly related to only satisfaction with the university, satisfaction with the university and intent to transfer were significantly and inversely related. It may be that perceptions with faculty behavior are directly related to satisfaction with the university and indirectly related to intent to remain and intent to transfer through satisfaction.

One limitation of the study is that the same sample was used for scale validation and testing relationships with outcome variables. However, tempering this limitation is that predictor and outcome variables were measured at different times, thus minimizing to some extent common method variance problems.

REFERENCES

Anderson, M. (1992). Imposters in the temple. New York: Simon & Schuster.

Appleby, D.C. (1990). Faculty and student perceptions of irritating behavior in the college classroom. The Journal of Staff, Program & Organization Development. 8, 41-46.

Ball, R. (2001). Practical difficulties with ethics. Public Management, 83(3), 2-5.

Branstetter, S.A. & M.M Handelsman. (2000). Graduate teaching assistants: Ethical training, beliefs, and practices. Ethics & Behavior, 10(1), 27-53.

Callahan, D. (1982). Should there be an academic code of ethics? Journal of Higher Education, 53, 335-344.

Gardner, D.G., L..L. Cummings, R.B. Dunham & J.L. Pierce. (1998). Single-Item versus multiple-item measurement scales: An empirical comparison. Educational & Psychological Measurement, 58(6), 898-915.

Garrett, T.M. & R.J. Klonoski. (1992). Business Ethics, 3rd Edition, Englewood Cliffs, NJ: Prentice-Hall.

Keith-Spiegel, P., G. Tabachnick & M. Allen. (1993). Ethics in academia: Students' views of professors' actions. Ethics & Behavior, 3, 149-162.

Kitchener, K.S. (1985). Ethical principles and ethical decisions in student affairs. In H.J. Canon & R.D. Brown (Eds.) New directions for student services: Applied ethics in student services. San Francisco, CA: Josey-Bass, 17-29.

Leo, J. (2002). Professors who see no evil. U. S. News & World Report, 133(3), 14.

Lipschutz, S.S. (1993). Enhancing success in doctoral education: From policy to practice. New Directions for Institutional Research, 80: 69-80.

Little, D. (1989). Are academicians by their very nature ethical, or should they be? Paper presented at the Annual Convention of the Virginia Community Colleges Association, Roanoke, VA.

Markie, P.J. (1994). A professor's duties: Ethical issues for college teaching. Lanhan, MD: Rowman & Littlefield.

Morgan, B.L. & A.J. Korschgen. (2001). The ethics of faculty behavior: Students' and professors' views. College Student Journal, 35(3), 418-432.

Murray, B. (2000). Professors' most grating behaviors. Monitor on Psychology, January, 56-67.

Nagy, M.S. (2002). Using a single-item approach to measure facet job satisfaction. Journal of Occupational & Organizational Psychology, 75(1), 77-87.

Nunnally, J. C. (1978). Psychometric theory. New York: McGraw-Hill.

Parker, C.E. (1997). Making retention work. Black Issues in Higher Education, 13(26), 120-128.

Podsakoff, P.M., & Organ, D. W. (1986). Self-reports in organizational research: Problems and prospects. Journal of Management, 12, 531-544.

Scarpello, V. & Campbell, J. P. (1983). Job satisfaction: Are all the parts there? Personnel Psychology, 36, 577-600.

Smith, R. (1996). Essential ethical considerations in education. Education, 117(1), 17-21.

Strike, K.A. & J.F. Soltis. (1992). The ethics of college teaching. New York: Teachers College Press.

Svinicki, M. (1994). Ethics in college teaching. In W.G. McKeachie (Ed.), Teaching tips: Strategies, research, and theory for college and university teachers. 9th Edition, Lexington, MA: D.C. Heath, 269-277.

Sykes, C. (1989). ProfScam: Professors and the demise of higher education. Washington, DC: Regnery Gateway, Inc.

Tabachnick, B.G., P.C. Keith-Spiegel & K.S. Pope. (1991). Ethics of teaching: Beliefs and behaviors of psychologies as educators. American Psychologist, 46, 506-515.

Wanous, J. P., Reichers, A. E. & Hudy, M. J. (1997). Overall job satisfaction: How good are single-item measures? Journal of Applied Psychology, 82, 247-252.

Whicker, M. & Kronenfeld, J. (1994). Dealing with ethical dilemmas on campus. Thousand Oaks, CA: Sage.

Zabihollah, R., R.C. Elmore & J.Z Szendi. (2001). Ethical behavior in higher educational institutions: The role of the code of conduct. Journal of Business Ethics, 30(2), 171-183.

Mel Schnake, Valdosta State University

William Fredenberger, Valdosta State University

Michael P. Dumler, Illinois State University

Table 1: Results of the Principal Components Analysis of Faculty
Ethical Behavior

 Factor 1 Factor 2
 Classroom Inappropriate
 Ethical Exchange
 Behavior

Giving lower grades to students .77 .00
who disagree with him/her (the
instructor).

Teaching class under the influence .75 .00
of alcohol or recreational drugs.

Not providing alternative teaching .65 .00
and testing procedures for students
who have learning disabilities.

Giving exams which do not reflect .73 .00
the material covered/discussed

Allowing how much he/she likes or .72 .23
dislikes a student to influence the
student's grade.

Including false or misleading .79 .00
information that may hurt the
student's chances when writing a
letter of recommendation for the
student.

Including material on a test that was .75 .00
not covered in the lectures or
assigned reading.

Requiring students to disclose .74 .23
highly personal information in a
group discussion or exercise (e.g.,
students who remain silent and don't
discuss the information are graded
down).

Asking small favors (such as a ride .00 .57
home) from students.

Accepting expensive gifts from .28 .70
students.

Borrowing money from students. .19 .82

Lending money to a student. .00 .78

Accepting inexpensive gifts from .00 .70
students.

Teaching a class without being .17 .00
adequately prepared that day.

Failure to maintain regularly .36 .17
scheduled office hours.

Frequently arriving several minutes .15 .00
late for class.

Frequently missing class without .43 .00
advance notice.

Not getting exams graded and .32 .00
returned until 4 weeks after the
exam was given.

Never learning any of the student's .00 .20
names in a relatively small class
(e.g., 20 to 30).

Being sexually attracted to a .00 .16
student.

Dating a student majoring in a field .00 .27
outside the professor's teaching
assignment and unlikely to ever
enroll in the professor's class.

Dating a student not currently .00 .31
enrolled in the professor's class.

Becoming sexually involved with a .00 .00
student after the course is
completed and grades are filed.

Hugging a student. .00 .00

Accepting a student's invitation to a .13 .28
party.

Selling goods (e.g., books, a car) to .00 .13
a student.

Hiring a student to work for him/her -.12 .20
(painting a house, baby-sit, etc.).

Beginning an on-going friendship .00 .20
with a student who is enrolled in the
professor's class.

Going to a bar with students after .33 .36
class.

Eigenvalue 7.96 4.85

% of Variance Explained 27.43 16.74

 Factor 3 Factor 4
 Relationships Dating
 Outside the
 Classroom

Giving lower grades to students .00 .00
who disagree with him/her (the
instructor).

Teaching class under the influence .00 .18
of alcohol or recreational drugs.

Not providing alternative teaching .00 .21
and testing procedures for students
who have learning disabilities.

Giving exams which do not reflect .00 .00
the material covered/discussed

Allowing how much he/she likes or .00 .00
dislikes a student to influence the
student's grade.

Including false or misleading .00 .00
information that may hurt the
student's chances when writing a
letter of recommendation for the
student.

Including material on a test that was .00 .00
not covered in the lectures or
assigned reading.

Requiring students to disclose .00 .00
highly personal information in a
group discussion or exercise (e.g.,
students who remain silent and don't
discuss the information are graded
down).

Asking small favors (such as a ride .38 .29
home) from students.

Accepting expensive gifts from .00 .21
students.

Borrowing money from students. .24 .11

Lending money to a student. .33 .14

Accepting inexpensive gifts from .23 .21
students.

Teaching a class without being .21 .00
adequately prepared that day.

Failure to maintain regularly .00 .00
scheduled office hours.

Frequently arriving several minutes .00 .00
late for class.

Frequently missing class without .00 .13
advance notice.

Not getting exams graded and .00 .14
returned until 4 weeks after the
exam was given.

Never learning any of the student's .00 .00
names in a relatively small class
(e.g., 20 to 30).

Being sexually attracted to a .21 .68
student.

Dating a student majoring in a field .16 .83
outside the professor's teaching
assignment and unlikely to ever
enroll in the professor's class.

Dating a student not currently .20 .82
enrolled in the professor's class.

Becoming sexually involved with a .26 .79
student after the course is
completed and grades are filed.

Hugging a student. .57 .33

Accepting a student's invitation to a .61 .16
party.

Selling goods (e.g., books, a car) to .74 .00
a student.

Hiring a student to work for him/her .73 .16
(painting a house, baby-sit, etc.).

Beginning an on-going friendship .66 .16
with a student who is enrolled in the
professor's class.

Going to a bar with students after .49 .26
class.

Eigenvalue 1.90 1.59

% of Variance Explained 6.56 5.46

 Factor 5
 Inappropriate
 Use of Time

Giving lower grades to students .12
who disagree with him/her (the
instructor).

Teaching class under the influence .18
of alcohol or recreational drugs.

Not providing alternative teaching .23
and testing procedures for students
who have learning disabilities.

Giving exams which do not reflect .23
the material covered/discussed

Allowing how much he/she likes or .00
dislikes a student to influence the
student's grade.

Including false or misleading .13
information that may hurt the
student's chances when writing a
letter of recommendation for the
student.

Including material on a test that was .18
not covered in the lectures or
assigned reading.

Requiring students to disclose .19
highly personal information in a
group discussion or exercise (e.g.,
students who remain silent and don't
discuss the information are graded
down).

Asking small favors (such as a ride .00
home) from students.

Accepting expensive gifts from .14
students.

Borrowing money from students. .00

Lending money to a student. .00

Accepting inexpensive gifts from .11
students.

Teaching a class without being .64
adequately prepared that day.

Failure to maintain regularly .56
scheduled office hours.

Frequently arriving several minutes .77
late for class.

Frequently missing class without .62
advance notice.

Not getting exams graded and .61
returned until 4 weeks after the
exam was given.

Never learning any of the student's .73
names in a relatively small class
(e.g., 20 to 30).

Being sexually attracted to a .12
student.

Dating a student majoring in a field .00
outside the professor's teaching
assignment and unlikely to ever
enroll in the professor's class.

Dating a student not currently .00
enrolled in the professor's class.

Becoming sexually involved with a .00
student after the course is
completed and grades are filed.

Hugging a student. .11

Accepting a student's invitation to a .00
party.

Selling goods (e.g., books, a car) to .17
a student.

Hiring a student to work for him/her .00
(painting a house, baby-sit, etc.).

Beginning an on-going friendship .13
with a student who is enrolled in the
professor's class.

Going to a bar with students after .00
class.

Eigenvalue 1.23

% of Variance Explained 4.25

Table 2: Correlations Among Faculty Ethical Behavior Dimensions

 Factor 1 Factor 2 Factor 3 Factor 4 Factor 5

Factor 1 .90 .40 ** .12 * .12 * .54 **
Factor 2 .86 .60 ** .49 ** .31 **
Factor 3 .79 .53 ** .21 **
Factor 4 .87 .19 **
Factor 5 .80