Faculty and Student Perceptions of the Seriousness, Prevalence, and Justification of Honor Code Violations at Connecticut College
Ishtiaq A. Mawla
Author note
Firstly, I would like to thank Dr. Valerie Wajda-Johnston at Tulane University for providing me with her questionnaires. In addition, I am indebted to Ms. Nova Seals, the Connecticut College archivist, and Dr. Marc Forster, for sharing valuable historical information on the college’s honor code. Moreover, Dr. Ann Sloan Devlin, instructor of Psychology 202, during the Spring 2012 term, shared her insights throughout the entire semester and has substantially developed my research skills. Finally, I thank Dr. Stuart Vyse for his advice on the statistical analyses. Correspondence concerning this article should be addressed to Ishtiaq A. Mawla, Department of Psychology, Connecticut College, New London, CT 06320. Email: [email protected]
Academic dishonesty and cheating are of critical concern in a college environment, particularly as they have dramatically increased over the past few decades (McCabe, Trevino, & Butterfield, 2001). The rate of cheating in 1940 was 20% which soared to about 80% in 1980 (Burns, Davis, Hoshino, & Miller, 1998). A review of 107 studies showed that 70.4% of college students had cheated to some extent, including 43.1% who had cheated on examinations, 40.9% who had cheated on homework assignments, and a further 47% who had plagiarized (Whitley, 1998). Dishonesty jeopardizes the pride that colleges and universities have of being unparalleled in ethical and moral standards (McCabe & Trevino, 1996). Extensive research has been done in this field; most studies have used self-report measures, where high levels of cheating were reported.
In a study where 200 randomly selected course catalogs from universities were analyzed, the researchers found basic themes in the universities’ dishonesty and cheating policies (Weaver, Davis, Look, Buzzanga, & Neal, 1991). One aspect of a typical policy expects students to voluntarily assume obligations of performance and behavior imposed by the institution relevant to its lawful missions and functions. Next, universities tend to assume that faculty members are responsible for clarification of honesty standards, and that it is the students’ responsibility to be familiar with the policies. In addition, most universities also describe what constitutes a dishonest act, including common acts such as plagiarism and direct cheating on an exam. Some universities also have an honor code system. Furthermore, institutions also tend to define the consequences of cheating and disciplinary actions that will be taken against a cheating student, which usually include academic probation, suspension, and/or expulsion. These penalties are typically imposed by a judicial or disciplinary committee that consists of faculty and/or deans.
There is no typical definition or demographic profile of students who engage in cheating-related behaviors. Age, gender, and academic achievement do not seem to predict cheating (Stephens & Nicholson, 2008). In addition, there seems to be no relationship between parental warmth or authoritativeness during childhood and attitudes toward academic dishonesty (Estep & Olsen, 2011). Classroom environment makes a difference in the amount of cheating; students who admitted cheating reported that their class was less personalized, less satisfying, and less task-oriented than did students who did not admit cheating (Pulvers & Diekhoff, 1999). Cheaters seem to have belief-behavior incongruity where their actions do not seem to match their beliefs. College students high in moral judgment cheated less often overall than did their low-reasoning counterparts, but they cheated just as often when the temptation to cheat became strong (Malinowski & Smith, 1985). On the other hand, non-cheaters, as compared to cheaters, are less likely to justify cheating, have higher rates of endorsing judicial restraints, and are more impacted by guilt (Vandehey, Diekhoff, & LaBeff, 2007).
One study (Jordan, 2001) showed that when cheating behavior was isolated to a single semester or term, 31% of participants cheated on a major exam or paper, but overall only 8.6% of students committed 75% of all acts of exam or paper cheating. This finding suggests that cheating behavior is spread out across a student’s academic career. Most studies of academic dishonesty are self-report, and these have also unveiled the excuses used by students to justify cheating behaviors. Students seem to compare themselves with their classmates and say that other students cheat more often than they do; thus, they believe that their cheating level is significantly lower than that of their peers (Del Carlo & Bodner, 2004). Students also seem to argue that cheating is an easy way to get whatever they want for a minimal amount of effort (Del Carlo & Bodner, 2004).
Deterrents against cheating include external and internal factors. External factors include fear of punishment, embarrassment, and fear of receiving a failing grade for an exam (Diekhoff et al., 1996). Internal factors consist of guilt and conscience, which seem to be the weakest deterrents against cheating (Diekhoff et al., 1996). Reduction of academic dishonesty primarily depends on faculty and institutional actions (Diekhoff et al., 1996).
Most faculty members are reluctant to report cheating, and this behavior seems to be fairly universal across numerous universities. However, the rate of reporting seems to be higher at institutions with an honor code (McCabe & Trevino, 1993). The hassle and time involved in each student case is usually the reason for reluctance (McCabe & Trevino, 1993). Some faculty members also seem to blame themselves for not making academic integrity clear to students (McCabe, 1993). Moreover, a survey of 127 professors showed that dealing with student cases of academic integrity was one of the most arduous aspects of their job, and there were five general reasons for not reporting such behavior (Keith-Spiegel, Tabachnick, Whitley, & Washburn, 1998). These five reasons consisted of avoidance of negative emotions (e.g., stress and anxiety during confrontation), denial that taking action is necessary (e.g., belief that students will eventually fail), fear of student retaliation, guilt (e.g., ruining a person’s career), and avoidance of a tedious process. Moreover, it was found that students who think or believe that faculty members are committed to the teaching and learning process were less likely to engage in academic dishonesty than were those who thought that their professors were not so engaged (McCabe & Trevino, 1996). However, with better understanding of an institution’s academic policies by both professors and students, dishonesty levels diminish and peer reporting of cheating situations increases (McCabe, Trevino, & Butterfield, 2002).
Honor codes are one of the most promising methods of promoting active student involvement in and responsibility for the maintenance of academic honesty (Simon et al., 2004). Honor code systems usually constitute a set of rules or principles for the students of an academic institution to follow. Under such environments, the faculty and administration assume that the students will behave responsibly in academic situations and place trust in their students.
Honor codes have been found to reduce academic dishonesty levels and increase willingness among faculty to report violations (McCabe & Trevino, 1993, 2002). Honor codes usually create clear definitions of cheating and their subsequent consequences (Jordan, 2001). Furthermore, honor code campuses tend to have a unique culture that supports the requirements of an honor code system. Traditional honor code systems often include unproctored examinations, the use of some form of written pledge in which students affirm that they have not cheated on a particular exam or assignment, the existence of a judicial or hearing body in which students play a major role (e.g., to inform other students about the purposes of the code, its major components, and the enforcement strategies), and the expectation of students to report any violations of the code they may observe (McCabe, Trevino, & Butterfield, 2002). One
study done at a non-honor code institution showed that merely creating an honor code and administering it to students were not sufficient to change the incidence of dishonesty, for students’ attitudes remained far too favorable with respect to cheating (Roig & Marks, 2006). Thus it may take years before the student culture changes. The aim of the current study was to explore academic dishonesty at an honor code institution – Connecticut College.
Connecticut College is a liberal arts institution with approximately 1,900 students, located in southern New England; it has a well-established honor code system (www.conncoll.edu). The college’s honor code has existed since 1911, the year the college was founded (Wade, 1987) and, in 1994, it was one of only 12 colleges in the United States to have an honor code system. The honor code is a bridge of honesty among the students, faculty, and administration, and it is assumed that every student at Connecticut College is mature enough to make his or her own decisions (Preston & Ammirati, 1991). Final examinations are unproctored and mostly self-scheduled. Every student is asked to write and sign the following pledge: “I promise to give nor receive any aid on this examination.” The college also employs a “lips sealed policy,” where students who have completed an examination are not allowed to disclose the content or even the level of difficulty of that examination to anyone before the end of the examination period (Connecticut College student handbook, 2012). The Honor Council is a student committee that deals with honor code violations. Any violation of academic integrity is dealt with by a temporary disciplinary probation, grade reduction for the exam or course, or even suspension, depending upon the context of the situation (Connecticut College student handbook, 2012).
Despite the existence of this honor code at Connecticut College, the faculty and administration realize that the system is not 100% effective, and some policies seem to confuse students (Woodsome, 1998). For example, some students do not believe in the system and claim that the honor code system is flawed, delusional, and even breeds cheaters on campus (Never, 1999; Ward, 2000; Woodsome, 1998). Theoretically, the honor code should allow students to play an active role in running the institution. However, a study done by the student government at Connecticut College showed that 9 out of 10 times, students do not turn in their peers for violating the honor code (Wade, 1987). Another study at Connecticut College showed that 43.8% of students thought that it was ideally more honorable to be loyal to a friend than to turn in a cheater, whereas 38.4% of students thought that turning in that friend was more honorable. Only 20.9% reported that they would actually turn in that person (Kreit, 2001). Another poll showed that 27% of students signed the honor code pledge knowing that they would break it (Kreit, 2001).
The present study serves to further previous research on academic dishonesty and cheating at Connecticut College. Not much research has been done on perceptions of academic dishonesty policy, and the literature on cheating mostly includes self-report studies, where questions about students’ own cheating behavior are usually asked, and this can be unreliable. A study at a large midwestern public university with no honor code (Wajda-Johnston, Handal, Brawer, & Fabricatore, 2001) explored three aspects, among many, of dishonesty perceptions: prevalence of academically dishonest behaviors, seriousness and severity of the violations, and the reasons used to justify dishonesty. Based on that earlier research, the following hypotheses were made:
Hypothesis 1: It was hypothesized that, within an honor code environment, faculty would estimate a lower level prevalence of cheating than would students.
Hypothesis 2: It was hypothesized that, within an honor code environment, faculty would report the severity of honor code violations to be higher than would students.
Hypothesis 3: It was hypothesized that, within an honor code environment, students would endorse more justifiable reasons for cheating than would faculty.
Method
Participants
There were two types of participants in this quasi-experimental study. The first group was composed of Connecticut College students from introductory psychology courses (PSY 101 and PSY 102) who participated in this study for course credit. One hundred and twenty-nine students took the survey, but only 118 students were included in the analyses; 11 student responses were excluded because they failed to respond to 75% or more of the questions or excluded one or more sections of the survey completely. Of the students who opted to report their sex, 89 were women and 27 were men. Fiftyseven students were freshmen, 38 were sophomores, 17 were juniors, and 4 were seniors. The mean age of students was 19.34; ages ranged from 18 to 22.
The second group of participants was composed of Connecticut College faculty across all departments. With a response rate of 75.3%, 61 out of 81 faculty members contacted took the survey. However, 44 faculty members were included in the analyses. Seventeen faculty responses were excluded because they failed to respond to 75% or more of the questions or excluded one or more sections of the survey completely. Out of the faculty members who opted to report their sex, 24 were women and 15 were men.
Materials
Questions for this study were adapted from a study that measured academic dishonesty at the graduate level (Wajda-Johnston et al., 2001), which used a modified version of the Cheating/Academic Dishonesty Survey (LaGrange, 1992). There were two questionnaires, one for faculty (with 112 items) and one for the students (with 106 items). Questions were modified to better represent undergraduate students and the Connecticut College academic environment. No information on internal consistency was reported by the Wajda-Johnston et al (2001) study.
The student survey examined perceived prevalence, perceived severity, and perceived justification of honor code violations at Connecticut College. As mentioned before, these have been explored at a large non-honor university at the graduate level (Wajda-Johnston et al., 2001). Questions about students’ own cheating behavior were omitted from the original survey because of the lack of reliability these self-report measures possess in sensitive matters such as cheating. Examples of questions from the survey include “Getting high grades is more important than learning” (this was rated on a 5 point scale from “Not at all dishonest” to “Severely dishonest”) and “Suppose a faculty member saw or knew of a student engaging in academically dishonest behaviors. Choose the ONE action that you REALISTICALLY EXPECT the faculty member to do” (this was a multiple choice question and the options included: A. Immediately confront the cheater, B. Warn the class that he/she is aware that dishonest behavior is occurring in the class but not mention the cheater’s name, C. Threaten the cheater that you will report him/her to the dean or honor council if the cheating does not stop, D. Tell the cheater you saw him/her cheating and you want it to stop, E. Mention to the cheater you saw him/her cheating, F. Ignore the cheating and do nothing, G. Tell another faculty member that the behavior is occurring, H. Immediately report the cheater to the dean or honor council, and I. Notify the dean or honor council that the behavior is occurring but not mention the cheater’s name). Demographic information, such as age, sex, and class year was collected at the end of the survey.
Students first responded to prevalence items (items 1 to 37), followed by severity items (items 38 to 74), and then justification items (items 75 to 92). Items 93 to 96 asked general questions on what participants perceive the actions students and faculty would ideally and realistically take when aware of or confronted by a dishonesty case. Items 97 and 98 asked how concerned students and faculty are on campus about cheating. Items 99 and 100 asked how concerned students and faculty on campus should be regarding cheating. Items 101 to 104 are demographic questions. Item 105 asked students whether they have answered all questions honestly.
The faculty survey also examined the same three topics of prevalence, severity, and justification. An example of a question from the justification section of the survey includes “Because so many other students cheat, you could be hurting your own grades if you didn’t cheat too” (this was rated on a 5 point scale of “Not a valid justification” to “Very strong”). Another example of a question from this survey was “Have you ever been reasonably certain that a student in one of your classes at Connecticut College was aware that cheating was occurring but chose to do nothing about it?” This was a “Yes” or “No” question.
On the faculty survey, items 1 to 37 examined prevalence levels, items 38 to 74 examined severity, and 77 to 95 examined justification (analogous to the student survey). Items 75 and 76 examined how many times the faculty participants have witnessed cheating and what percentage of students do engage in academically dishonest behaviors. Items 96 to 99 included what students and faculty would realistically and ideally do in cases of
dishonesty. Items 100 and 101 explored whether faculty encountered situations where students and other faculty members have not taken any actions against cheating. Items 102 to 104 examined how concerned faculty and students are about dishonesty levels. Items 105 and 106 examined how concerned about cheating they should really be.
A reliability analysis was conducted on the modified scale, and it was found to be highly reliable overall, Cronbach’s α = .93. The separate subscales also were found to be reliable: prevalence subscale, Cronbach’s α = .97; severity subscale, Cronbach’s α = .97; justification subscale, Cronbach’s α = .95.
Procedure
Student signup sheets were posted on the second floor of Bill Hall at Connecticut College on a bulletin board next to other signup sheets for other ongoing research projects. Participants were PSY 101 and PSY 102 students. The survey was administered online through Qualtrics TM and the Connecticut College Survey Monkey TM account. URLs to the survey were emailed to each participant. The online survey required the students to fill out the informed consent form. A debriefing form appeared at the end of the survey, and participants were asked to print it out to obtain course credit for research participation.
Faculty surveys were sent out through email using a modified Dillman (2007) survey method. The first email directly has the URL to the questionnaire, and it requested participation and explained the importance of the study. Two subsequent weekly reminders were sent after the first email. Every survey had the informed consent form at the beginning and the debriefing form at the end.
Results
The purpose of this study was to examine the perceptions of prevalence and severity of academic dishonesty and of justifications given for academically dishonest behaviors across students and faculty at Connecticut College. To distill the ratings of prevalence, severity, and justification, the dimensions were identified through factor analysis, a process used to discover patterns of relationships among items, and create independently grouped subsets, which usually have an ascribed meaning and convey certain information about that particular subset. In this case, the grouping occurs based on the rating each cheating-related behavior received. The factor analysis method employed here involved a Principal Components Analysis extraction method and a Varimax with Kaiser Normalization rotation method, and this was conducted separately for each of the three components (prevalence, severity, and justification).
For prevalence, the initial solution produced five factors with eigenvalues greater than 1 (the rotation converged in eight iterations). Sixteen items did not meet the minimum loading criterion of .40, or loaded at that level or higher on more than one factor. The fourth and fifth factors were not well defined, with very few items per factor. Consequently, a subsequent analysis was done to extract and rotate three factors (the rotation converged in 5 iterations). The third factor was not used because it only contained two items, and it was not found to be prominent after examination of the scree plot. Fourteen of the original 37 items clustered on two factors that were labeled Category 1 (normal) and Category 2 (extreme), which accounted for 56.39% of the variance. The first factor, Category 1 (normal), accounted for 32.79% of the variance and contained 8 items. Normal dishonest behaviors grouped into this category. The second factor, Category 2 (extreme) accounted for 23.60% of the variance and contained 6 items. Extremely dishonest behaviors, including those involving monetary and sexual compensation, grouped into this category. Table 1 presents the items and factor loadings.
For severity, the initial solution produced three factors with eigenvalues greater than 1 (the rotation converged in four iterations). Twelve items did not load greater than .40 or loaded at that level or higher on more than one factor. The third factor was not very well defined, as it had one item. Consequently, a subsequent analysis was done to extract and rotate three factors again (the rotation converged in four iterations). The third factor was not used because there was just one item and it was not found to be prominent after examination of the scree plot. Twenty-three of the original 37 items clustered on two factors that were labeled Category 1 (extreme) and Category 2 (normal), which accounted for 62.32% of the variance. The first factor, Category 1 (extreme), accounted for 35.11% of the variance and contained 13 items. As with prevalence, extremely dishonest behaviors, including those involving monetary and sexual compensation, grouped into this category. The second factor, Category 2 (normal) accounted for 27.21% of the variance and contained 10 items. This factor contained “normal” academically dishonest behaviors. Table 2 presents the items and the factor loadings.
For justification, the initial analysis produced three factors with eigenvalues greater than 1 (the rotation converged in five iterations). Two items did not load greater than .40 or loaded at that level or higher on more than one factor. The third factor was not very well defined, as it had just one item. Consequently, a subsequent analysis was done to extract and rotate three factors again (the rotation converged in five iterations). The third factor was not used because there was just one item, which was not found to be prominent after examination of the scree plot. Sixteen of the 19 original items clustered on two factors that were labeled “grade/career-oriented” and “situational,” which accounted for 60.87% of variance. The first factor, grade/career-oriented accounted for 31.59% of the variance and contained 8 items. This factor contained items pertaining to student beliefs regarding the importance of good grades, regardless of how they are achieved. It also consisted of items that concern the importance of high grades for future career plans. The second factor, situational, accounted for 29.28% of the variance and also contained 8 items. This factor consisted of items concerning the education and examination environment and the punishments involved. Table 3 presents the items and the factor loadings.
Items that loaded on each factor were summed, divided by the number of items in each factor, and used in subsequent analyses. For each of the items in a given factor for prevalence, the higher the value, the more prevalent academic dishonesty is perceived to be (e.g., 1 = 0% and 11 = 91-100%). For each of the items in a given factor for severity, the higher the value, the more severe academic dishonesty is perceived to be (e.g., 1 = Not at all dishonest and 5 =Severely dishonest). For each of the items in a given factor for justification, the higher the value, the less justifiable the behavior is perceived to be (e.g., 1 = Not a valid justification to 5 = Very strong). Table 4 presents the overall means and standard deviations of the means of the factors in each factor (prevalence, severity, and justification).
To evaluate the first hypothesis that, within an honor code environment, faculty would estimate a lower prevalence of cheating than would students, a one-way multivariate analysis of variance (MANOVA) was conducted. The two levels of type (students and faculty) served as the fixed factor, and the means of the two factors obtained through factor analysis, Category 1 (normal) and Category 2 (extreme), served as the dependent variables. The analysis was significant, Wilks’s λ = .830, F(2, 156) = 10.64, p <.001. Univariate analyses of variance revealed significant differences for Category 1 (normal), F(1, 158) = 9.22, p = .003, but no significant differences were found for Category 2 (extreme), F(1, 158) = 2.71, p = .102, between students and faculty. In the normal category, faculty judged the prevalence of cheating to be less prevalent than did students. Means and standard deviations of the prevalence ratings, for both students and faculty, and the results of the univariate analyses are presented in Table 5.
To evaluate the next hypothesis that, within an honor code environment, faculty would report the severity of honor code violations to be greater than would students, a one-way MANOVA was conducted, with the two levels of type (students and faculty) as the fixed factor, and the means of the two factors obtained through factor analysis distillation, Category 1 (extreme) and Category 2 (normal), as the dependent variables. The analysis was significant, Wilks’s λ = .772, F(2, 159) = 23.52, p < .001. Univariate analyses of variance revealed significant differences for Category 1 (extreme), F(1, 160) = 7.14, p =.008, and Category 2 (normal), F(1, 160) = 46.16, p < .001, between students and faculty. In both categories, faculty rated the behaviors to be more severe than did students. Means and standard deviations of the severity ratings, for both students and faculty, and the results of the univariate analysis are presented in Table 6.
To evaluate the final hypothesis that, within an honor code environment, students would endorse more justifiable reasons for cheating than would faculty, a one-way MANOVA was conducted, with two levels of type (students and faculty) as the fixed factor, and the means of the two factors obtained through factor analysis distillation, grade/career-oriented and situational, as the dependent variables. The analysis was significant, Wilks’s λ =.670, F(2, 158) = 38.83, p < .001. Univariate analyses of variance revealed significant differences for grade/career-oriented, F(1, 159) = 68.67, p < .001, and situational, F(1, 159) = 14.10, p < .001, between students and faculty. Students provided higher ratings on the justification items (of both categories) than did faculty. Means and standard deviations of the justification ratings, for both students and faculty, and the results of the univariate analyses, are present in Table 7.
In addition, further analysis was done on the student data to examine whether there were differences between women and men. A one-way MANOVA with two levels of sex (women and men) as the fixed factor and the factors for justification as the dependent variables approached significance,Wilks’s λ = .950, F(2, 113) = 2.95, p = .056. An additional one-way MANOVA with two levels of sex (women and men) as the fixed factor and the factors for severity as the dependent variables also approached significance, Wilks’s λ =.952, F(2, 113) = 2.83, p = .063.
Certain questions that produced categorical data were also asked to both students and faculty near the end of the survey. One of these questions was “How concerned do the following groups at Connecticut College appear to be about cheating/academic dishonesty?” Among students, 42.7% thought that faculty care a great deal about academic dishonesty and 9.5% thought that students care a great deal about academic dishonesty; while among faculty, 41.5% thought that they themselves care a great deal about academic dishonesty, only 4.9% thought that students care a great deal about academic dishonesty, and 29.3% thought that other faculty care a great deal about academic dishonesty.
In addition, among students, 29.1% thought that faculty care a good deal about academic dishonesty and 33.6% thought that students care a good deal about academic dishonesty; while among faculty, 34.1% thought that they themselves care a good deal about academic dishonesty, 26.8% thought that students care a good deal about academic dishonesty, and 43.9% thought that other faculty care a good deal about academic dishonesty.
Moreover, among students, 17.1% thought that faculty somewhat care about academic dishonesty and 33.6% thought that students somewhat care about academic dishonesty; while among faculty, 19.5% thought that they themselves somewhat care about academic dishonesty, 46.3% thought that students somewhat care about academic dishonesty, and 17.1% thought that other faculty somewhat care about academic dishonesty.
Furthermore, among students, 8.5% thought that faculty minimally care about academic dishonesty and 18.1% thought that students minimally care about academic dishonesty; while among faculty, 2.4% thought that they themselves minimally care about academic dishonesty, 22% thought that students minimally care about academic dishonesty, and 7.3% thought that other faculty minimally care about academic dishonesty.
Finally, among students, 2.6% thought that faculty did not care about academic dishonesty at all and 5.2% thought that students did not care about academic dishonesty at all; while among faculty, 2.4% thought that they themselves did not care about academic dishonesty at all, no one thought that students did not care about academic dishonesty at all, and 2.4% thought that other faculty did not care about academic dishonesty at all.
Another question that produced categorical data was “Do you think the following groups at Connecticut College should be more concerned about cheating than they currently appear to be?” Among students, 6.8% thought that faculty should be much more concerned about academic dishonesty and 5.1% thought that students should be much more concerned about academic dishonesty; while among faculty, no one thought that they themselves should be much more concerned about academic dishonesty, 40% thought that students should be much more concerned about academic dishonesty, and 10.3% thought that other faculty should be much more concerned about academic dishonesty.
In addition, among students, 38.5% thought that faculty should be more concerned about academic dishonesty and 12% thought that students should be more concerned about academic dishonesty; while among faculty, 20% thought that they themselves should be more concerned about academic dishonesty, 42.5% thought that students should be more concerned about academic dishonesty, and 33.3% thought that other faculty should be more concerned about academic dishonesty.
Moreover, among students, 47.9% thought that faculty should not change their level of concern about academic dishonesty and 77.8% thought that students should not change their level of concern about academic dishonesty; while among faculty, 75% thought that they themselves should not change their level of concern about academic dishonesty, 15% thought that students should not change their level of concern about academic dishonesty, and 53.8% thought that other faculty should not change their level of concern about academic dishonesty.
Furthermore, among students, 3.4% thought that faculty should be less concerned about academic dishonesty and 2.6% thought that students should be less concerned about academic dishonesty; while among faculty, 5% thought that they themselves should be less concerned about academic dishonesty, 2.5% thought that students should be less concerned about academic dishonesty, and 2.6% thought that other faculty should be less concerned about academic dishonesty.
Finally, among students, 3.4% thought that faculty should be much less concerned about academic dishonesty and no one thought that students should be much less concerned about academic dishonesty; while among faculty, no one thought that they themselves, students, or other faculty should be much less concerned about academic dishonesty.
Over 85% of faculty members (n = 37) said that they address academic dishonesty on the first day of classes whereas 14% (n = 6) said that they do not. Over 80% of faculty members (n = 36) said that they address cheating in their course syllabi, whereas 16% (n = 7) said that they do not.
Discussion
Do students and faculty perceive academic infractions differently in the context of an honor code? Results indicate that compared to students, faculty members tend to estimate a lower level of prevalence of general academically dishonest behaviors. However, this was not the case with extreme cases of dishonesty: Faculty and students rated them similarly. Extreme cases of dishonesty were defined (through factor analysis) as those behaviors that included some form of monetary or sexual compensation, such as in cases where a student took an examination for another student. Both students and faculty did not perceive extreme cases to be very prevalent. These results of the present study are supported by those of Brimble and Stevenson-Clarke (2005). Wajda-Johnston et al. (2001) found significant differences on certain prevalence ratings between students and faculty, but concluded that they were not meaningful because, despite the significance difference, the means for both faculty and students were found to be in the same percentage range.
However, the current study revealed that the prevalence ratings, in addition to statistical significance, are meaningful differences. Students generally perceive prevalence of cheating to be in the 21 - 30% range, whereas faculty perceive it to be in the 11 - 20% range. A reason behind this difference may be that students are more likely than faculty to witness academic dishonesty first hand in situations such as unproctored examinations or take-home assignments. It also may be that faculty are not usually physically present in examination situations or dormitory and library environments, and that they place their full trust in students based on the honor code; consequently, they may not really be aware of much cheating and dishonesty. The ratings of prevalence found in this study are higher than those found in the previous study done at a graduate level in a midwestern, religiously affiliated university (Wajda-Johnston et al., 2001), which may suggest that dishonesty at the graduate level is perceived is perceived to be less prevalent that it is perceived to be at the undergraduate level. Moreover, students in the Wajda-Johnston et al. (2001) study may be more aware of the ethical/moral dimensions of cheating, because of the religious affiliation of the university.
Results also indicate that faculty reported honor code violations to be more severe than did students. This difference was present in both normal and extreme cases of dishonest behaviors. Extreme cases of dishonesty were rated more harshly in terms of severity than were normal cases of dishonesty. The mean ratings in both extreme and normal groups were very high across both groups of faculty and students. There were no ratings below 3.6 (on a scale of 5, where 5 is the most severe), and scores were rated “moderately dishonest” or “severely dishonest,” which shows that, despite their differences, both students and faculty in an honor code environment perceive academically dishonest behaviors to be quite severe and morally incorrect. However, the means (and the low standard deviations) did not seem to make a practically meaningful difference, because, after score conversion, both students and faculty in the extreme category rated the behaviors to be “severely dishonest,” and both students and faculty in the normal category rated the behaviors to be “moderately dishonest.” This result is supported by previous research (Wajda-Johnston et al., 2001). This result also seemed to be similar to another study that concerned severity perceptions (Sims, 1995), where it was found that faculty generally consider dishonest behaviors to be more severe than do students. A possible reason for faculty’s harsh judgment of academic dishonesty is because it is their duty to uphold the honor code, and they act as the prime administrators and implementers of the system. Also, faculty view academic work in a professional way.
For justification, students seem to endorse more reasons for cheating that did faculty. This was the case in both grade/career-oriented and Situational factors. Very few items were rated as “moderately strong” or “very strong” in terms of justification, which shows that, in general, both students and faculty are less willing to support reasons given as justifying academic dishonesty. The difference between faculty and students on grade/career-oriented ratings was found to be practically meaningful; faculty rated these items from “not a valid justification” to “very weak,” whereas students rated them from “very weak” to “slightly weak.” However, the difference between faculty and students on situational items was not found to be practically meaningful, and both students and faculty rated them from “not a valid justification” to “very weak,” which was slightly different from previous research (Wajda-Johnston et al., 2001) and showed that most of the items did not produce meaningful differences between students and faculty. The distilled justification factors produced are consistent with previous findings, which showed that competition for grades, insufficient study time, and desire for success are some reasons why students cheat (Baird, 1980; Stevens & Stevens, 1987). A possible reason why students endorse more academically dishonest behaviors as justifiable could be their stressful and competitive academic environment, which constantly compels them to strive for the best grades and academic opportunities. It could also be that they are more likely to have recently engaged in those behaviors, and therefore find them justifiable to relieve guilt.
Results from this study indicate that college administrators should reevaluate the honor code system as there is a clear discrepancy in how faculty and students perceive the honor code. It is alarming that students see academic dishonesty to be more justifiable and less severe than the faculty do.
Limitations and Practical implications
The current study has several limitations. The survey had certain questions that were aimed to examine cheating behavior in laboratory classes. However, such classes are mostly taught by faculty in the sciences. These questions, thus, were not applicable across all faculty. Moreover, not all professors at Connecticut College administer examinations or papers. Some choose one over the other, some administer both, and faculty in the arts assign other types of work. The faculty survey should have accounted for these particular cases. The number of participants in the faculty sample was not very high (N = 44). The response rate of faculty was fairly high, but it would have helped the power of the sample if more faculty had taken the survey. The student participants were skewed in terms of gender; 76.7% of the student sample consisted of women. Differences in perception of academic dishonesty between women and men in the student data were also examined, and the multivariate analysis approached significance (p = .056 for justification and p = .063 for severity). Significant results might have been revealed if there were a larger number of male participants. The student data also were skewed in terms of class year, as most introductory psychology students are freshmen and sophomores. Of the 116 students who chose to report their class year, 17 were juniors and only 4 were seniors. Being at the college and thus under the honor code for a longer time could affect beliefs about the system. The class distribution could be improved by conducting the survey campus-wide, instead of just restricting the participant pool to introductory psychology students.
It is also unclear whether the perceptions of prevalence given by students are reflective of their own cheating behaviors, as self-report of own behaviors was completely omitted from the study. Generally, self-report measures of cheating have revealed higher prevalence (McCabe, Butterfield, & Trevino, 2006).
Future Directions
The survey should be administered again to the entire campus community to achieve a wider participant pool. In addition to dealing with academic dishonesty, the honor code at Connecticut College deals with social and community infractions (Rodgers, 2001) and it could be valuable if violations of social aspects of the honor code were also examined. Comparisons could be made between perceptions of severity of academic honor code violation versus social honor code violations. In addition, borderline cases of academic dishonesty could be examined. These can be defined where there are no clear guidelines or rules in the honor code policy on whether that particular behavior constitutes as academic dishonesty or not. For this study, self-report measures of cheating were omitted because of the sensitivity of the issue. In future research, such self-report measures could actually be administered to look for differences between cheaters and non-cheaters in terms of their perceptions of prevalence, severity, and justification of academic dishonesty.
References
Baird, J. S. (1980). Current trends in college cheating. Psychology in the Schools, 17, 515-522. doi: 10.1002/1520-6807(198010)17:4<515::AIDPITS2310170417>3.0.CO;2-3
Brimble, M., & Stevenson-Clarke, P. (2005). Perceptions of the prevalence and seriousness of academic dishonesty in Australian universities. Australian Educational Researcher, 32, 19-44. doi:10.1007/BF03216825
Burns, S. R., Davis, S. F., Hoshino, J., & Miller, R. L. (1998). Academic dishonesty: A delineation of cross-cultural patterns. College Student Journal, 32, 590-596. doi:10.1080/10508420701713030
Connecticut College. (2012). Connecticut College student handbook. Retrieved from http://www.conncoll.edu/cw40/c/document_library/get_file?uuid=8eea2e03-4d6a-4d33-92bf-5f6abc8a3a47&groupId=10143
Del-Carlo, D. I., & Bodner, G. M. (2004). Students' perceptions of academic dishonesty in the chemistry classroom laboratory. Journal of Research in Science Teaching, 41, 47-64. doi:10.1002/tea.10124
Diekhoff, G. M., LaBeff, E. E., Clark, R. E., Williams, L. E., Francis, B., & Haines, V. J. (1996). College cheating: Ten years later. Research in Higher Education, 37, 487-502. doi:10.1007/BF01730111
Dillman, D. A. (2007). Mail and internet surveys: The tailored design method. Hoboken, New Jersey: Wiley Estep, H. M., & Olson, J. N. (2011). Parenting style, academic dishonesty, and infidelity in college students. College Student Journal, 45, 830-838. doi: 10.1007/s 10551-007-9576-0
Jordan, A. E. (2001). College student cheating: The role of motivation, perceived norms, attitudes, and knowledge of institutional policy. Ethics & Behavior, 11, 233-247. doi:10.1207/S15327019EB1103_3
Keith-Spiegel, P., Tabachnick, B. G., Whitley, B. E., & Washburn, J. (1998). Why professors ignore cheating: Opinions of a national sample of psychology instructors. Ethics & Behavior, 8, 215-227. doi:10.1207/s15327019eb0803_3
Kreit, B. (2001, September 21). Survey shows that friendships come before conscience. College Voice, p. 1.
LaGrange, M. V. (1992). Student, faculty, and administrator perceptions of academic dishonesty. Unpublished master’s thesis, Saint Louis University, Saint Louis, Missouri.
Malinowski, C. I., & Smith, C. P. (1985). Moral reasoning and moral conduct: An investigation prompted by Kohlberg's theory. Journal of Personality and Social Psychology, 49, 1016-1027. doi:10.1037%2F%2F0022-3514.49.4.1016
McCabe, D. L. (1993). Faculty responses to academic dishonesty: The influence Perceptions of Academic Dishonesty 70 of student honor codes. Research in Higher Education. 34, 647-658. doi:10.1007/BF00991924
McCabe, D. L., Butterfield, K. D., & Trevino, L. K. (2006). Academic dishonesty in graduate business programs: Prevalence, causes, and proposed action. Academy of Management Learning & Education, 5, 294-305. doi:10.5465/AMLE.2006.22697018
McCabe, D. L., & Trevino, L. K. (1993). Academic dishonesty: Honor codes and other contextual influences. Journal of Higher Education, 64, 522-538.
McCabe, D. L., & Trevino, L. K. (1996). What we know about cheating in college: Longitudinal trends and recent developments. Change, 28, 28–32.
McCabe, D. L., & Trevino, L. K., & Butterfield, K. D. (2001). Cheating in academic institutions: A decade of research. Ethics and Behavior, 11, 219-232. doi:10.1207/S15327019EB1103_2
McCabe, D. L. & Trevino, L. K., & Butterfield, K. D. (2002). Honor codes and other contextual influences on academic integrity: A replication and extension to modified honor code settings. Research in Higher Education, 43, 357-378. doi:10.1023/A:1014893102151
Never, B. (1999, February 26). Problems lie with the honor code, not confidentiality. College Voice, p. 2.
Preston, T., & Ammirati, J. (1991, September 3). Our school
and our honor code. College Voice, p. 2.
Pulvers, K., & Diekhoff, G. M. (1999). The relationship between academic honesty and college classroom environment. Research in Higher Education, 487-498.
Rodgers, J. (2001, November 9). J-board statistics reveal more male honor code violations. College Voice, p. 1.
Roig, M., & Marks, A. (2006). Attitudes toward cheating before and after the implementation of a modified honor code: A case study. Ethics & Behavior, 16, 163-171. doi:10.1207/s15327019eb1602_6
Simon, C. A., Carr, J. R., McCullough, S. M., Morgan, S. J., Oleson, T., & Ressel, M. (2004). Gender, student perceptions, institutional commitments, and academic dishonesty: Who reports in academic dishonesty cases? Assessment and Evaluation in Higher Education, 29, 75-90. doi:10.1080/026029303200015817
Sims, R. I. (1995). The severity of academic dishonesty: A comparison of faculty and student views. Psychology in the Schools, 32, 233-238. doi:10.1002/1520-6807(199507)32:3<233::AIDPITS2310320311>3.0.CO;2-H
Stephens, J. M., & Nicholson, H. (2008). Cases of incongruity: Exploring the divide between adolescents' beliefs and behavior related to academic dishonesty. Educational Studies, 34, 361-376. doi:10.1080/03055690802257127
Stevens, G. E., & Stevens, F. W. (1987). Ethical inclinations of tomorrow’s managers revisited: How and why students cheat. Journal of Education for Business, 63, 24-29. doi:10.1080/08832323.1987.10117269
Vandehey, M. A., Diekhoff, G. M., & LaBeff, E. E. (2007). College cheating: A twenty-year follow-up and the addition of an honor code. Journal of College Student Development, 48, 468-480. doi:10.1353/csd.2007.0043
Wade, P. (1987, November 3). The honor code analyzed: Is it only an ideal? College Voice, p. 4.
Wajda-Johnston, V. A., Handal, P. J., Brawer, P. A., & Fabricatore, A. N. (2001). Academic dishonesty at the graduate level. Ethics & Behavior, 11, 287-305. doi:10.1207/S15327019EB1103_7
Ward, C. (2000, May 8). Honor code review could spell the end for self scheduled exams. College Voice, p. 1.
Weaver, K. A., Davis, S. F., Look, C., Buzzanga, V. L., & Neal, L. (1991). Examining academic dishonesty policies. College Student Journal, 25, 302-305.
Whitley, B. E. (1998). Factors associated with cheating among college students: A review. Research in Higher Education, 39, 235-274. doi:10.1023/A:1018724900565
Woodsome, K. (1998, October 23). Honor code confuses Connecticut college students. College Voice, p. 1
Firstly, I would like to thank Dr. Valerie Wajda-Johnston at Tulane University for providing me with her questionnaires. In addition, I am indebted to Ms. Nova Seals, the Connecticut College archivist, and Dr. Marc Forster, for sharing valuable historical information on the college’s honor code. Moreover, Dr. Ann Sloan Devlin, instructor of Psychology 202, during the Spring 2012 term, shared her insights throughout the entire semester and has substantially developed my research skills. Finally, I thank Dr. Stuart Vyse for his advice on the statistical analyses. Correspondence concerning this article should be addressed to Ishtiaq A. Mawla, Department of Psychology, Connecticut College, New London, CT 06320. Email: [email protected]
Academic dishonesty and cheating are of critical concern in a college environment, particularly as they have dramatically increased over the past few decades (McCabe, Trevino, & Butterfield, 2001). The rate of cheating in 1940 was 20% which soared to about 80% in 1980 (Burns, Davis, Hoshino, & Miller, 1998). A review of 107 studies showed that 70.4% of college students had cheated to some extent, including 43.1% who had cheated on examinations, 40.9% who had cheated on homework assignments, and a further 47% who had plagiarized (Whitley, 1998). Dishonesty jeopardizes the pride that colleges and universities have of being unparalleled in ethical and moral standards (McCabe & Trevino, 1996). Extensive research has been done in this field; most studies have used self-report measures, where high levels of cheating were reported.
In a study where 200 randomly selected course catalogs from universities were analyzed, the researchers found basic themes in the universities’ dishonesty and cheating policies (Weaver, Davis, Look, Buzzanga, & Neal, 1991). One aspect of a typical policy expects students to voluntarily assume obligations of performance and behavior imposed by the institution relevant to its lawful missions and functions. Next, universities tend to assume that faculty members are responsible for clarification of honesty standards, and that it is the students’ responsibility to be familiar with the policies. In addition, most universities also describe what constitutes a dishonest act, including common acts such as plagiarism and direct cheating on an exam. Some universities also have an honor code system. Furthermore, institutions also tend to define the consequences of cheating and disciplinary actions that will be taken against a cheating student, which usually include academic probation, suspension, and/or expulsion. These penalties are typically imposed by a judicial or disciplinary committee that consists of faculty and/or deans.
There is no typical definition or demographic profile of students who engage in cheating-related behaviors. Age, gender, and academic achievement do not seem to predict cheating (Stephens & Nicholson, 2008). In addition, there seems to be no relationship between parental warmth or authoritativeness during childhood and attitudes toward academic dishonesty (Estep & Olsen, 2011). Classroom environment makes a difference in the amount of cheating; students who admitted cheating reported that their class was less personalized, less satisfying, and less task-oriented than did students who did not admit cheating (Pulvers & Diekhoff, 1999). Cheaters seem to have belief-behavior incongruity where their actions do not seem to match their beliefs. College students high in moral judgment cheated less often overall than did their low-reasoning counterparts, but they cheated just as often when the temptation to cheat became strong (Malinowski & Smith, 1985). On the other hand, non-cheaters, as compared to cheaters, are less likely to justify cheating, have higher rates of endorsing judicial restraints, and are more impacted by guilt (Vandehey, Diekhoff, & LaBeff, 2007).
One study (Jordan, 2001) showed that when cheating behavior was isolated to a single semester or term, 31% of participants cheated on a major exam or paper, but overall only 8.6% of students committed 75% of all acts of exam or paper cheating. This finding suggests that cheating behavior is spread out across a student’s academic career. Most studies of academic dishonesty are self-report, and these have also unveiled the excuses used by students to justify cheating behaviors. Students seem to compare themselves with their classmates and say that other students cheat more often than they do; thus, they believe that their cheating level is significantly lower than that of their peers (Del Carlo & Bodner, 2004). Students also seem to argue that cheating is an easy way to get whatever they want for a minimal amount of effort (Del Carlo & Bodner, 2004).
Deterrents against cheating include external and internal factors. External factors include fear of punishment, embarrassment, and fear of receiving a failing grade for an exam (Diekhoff et al., 1996). Internal factors consist of guilt and conscience, which seem to be the weakest deterrents against cheating (Diekhoff et al., 1996). Reduction of academic dishonesty primarily depends on faculty and institutional actions (Diekhoff et al., 1996).
Most faculty members are reluctant to report cheating, and this behavior seems to be fairly universal across numerous universities. However, the rate of reporting seems to be higher at institutions with an honor code (McCabe & Trevino, 1993). The hassle and time involved in each student case is usually the reason for reluctance (McCabe & Trevino, 1993). Some faculty members also seem to blame themselves for not making academic integrity clear to students (McCabe, 1993). Moreover, a survey of 127 professors showed that dealing with student cases of academic integrity was one of the most arduous aspects of their job, and there were five general reasons for not reporting such behavior (Keith-Spiegel, Tabachnick, Whitley, & Washburn, 1998). These five reasons consisted of avoidance of negative emotions (e.g., stress and anxiety during confrontation), denial that taking action is necessary (e.g., belief that students will eventually fail), fear of student retaliation, guilt (e.g., ruining a person’s career), and avoidance of a tedious process. Moreover, it was found that students who think or believe that faculty members are committed to the teaching and learning process were less likely to engage in academic dishonesty than were those who thought that their professors were not so engaged (McCabe & Trevino, 1996). However, with better understanding of an institution’s academic policies by both professors and students, dishonesty levels diminish and peer reporting of cheating situations increases (McCabe, Trevino, & Butterfield, 2002).
Honor codes are one of the most promising methods of promoting active student involvement in and responsibility for the maintenance of academic honesty (Simon et al., 2004). Honor code systems usually constitute a set of rules or principles for the students of an academic institution to follow. Under such environments, the faculty and administration assume that the students will behave responsibly in academic situations and place trust in their students.
Honor codes have been found to reduce academic dishonesty levels and increase willingness among faculty to report violations (McCabe & Trevino, 1993, 2002). Honor codes usually create clear definitions of cheating and their subsequent consequences (Jordan, 2001). Furthermore, honor code campuses tend to have a unique culture that supports the requirements of an honor code system. Traditional honor code systems often include unproctored examinations, the use of some form of written pledge in which students affirm that they have not cheated on a particular exam or assignment, the existence of a judicial or hearing body in which students play a major role (e.g., to inform other students about the purposes of the code, its major components, and the enforcement strategies), and the expectation of students to report any violations of the code they may observe (McCabe, Trevino, & Butterfield, 2002). One
study done at a non-honor code institution showed that merely creating an honor code and administering it to students were not sufficient to change the incidence of dishonesty, for students’ attitudes remained far too favorable with respect to cheating (Roig & Marks, 2006). Thus it may take years before the student culture changes. The aim of the current study was to explore academic dishonesty at an honor code institution – Connecticut College.
Connecticut College is a liberal arts institution with approximately 1,900 students, located in southern New England; it has a well-established honor code system (www.conncoll.edu). The college’s honor code has existed since 1911, the year the college was founded (Wade, 1987) and, in 1994, it was one of only 12 colleges in the United States to have an honor code system. The honor code is a bridge of honesty among the students, faculty, and administration, and it is assumed that every student at Connecticut College is mature enough to make his or her own decisions (Preston & Ammirati, 1991). Final examinations are unproctored and mostly self-scheduled. Every student is asked to write and sign the following pledge: “I promise to give nor receive any aid on this examination.” The college also employs a “lips sealed policy,” where students who have completed an examination are not allowed to disclose the content or even the level of difficulty of that examination to anyone before the end of the examination period (Connecticut College student handbook, 2012). The Honor Council is a student committee that deals with honor code violations. Any violation of academic integrity is dealt with by a temporary disciplinary probation, grade reduction for the exam or course, or even suspension, depending upon the context of the situation (Connecticut College student handbook, 2012).
Despite the existence of this honor code at Connecticut College, the faculty and administration realize that the system is not 100% effective, and some policies seem to confuse students (Woodsome, 1998). For example, some students do not believe in the system and claim that the honor code system is flawed, delusional, and even breeds cheaters on campus (Never, 1999; Ward, 2000; Woodsome, 1998). Theoretically, the honor code should allow students to play an active role in running the institution. However, a study done by the student government at Connecticut College showed that 9 out of 10 times, students do not turn in their peers for violating the honor code (Wade, 1987). Another study at Connecticut College showed that 43.8% of students thought that it was ideally more honorable to be loyal to a friend than to turn in a cheater, whereas 38.4% of students thought that turning in that friend was more honorable. Only 20.9% reported that they would actually turn in that person (Kreit, 2001). Another poll showed that 27% of students signed the honor code pledge knowing that they would break it (Kreit, 2001).
The present study serves to further previous research on academic dishonesty and cheating at Connecticut College. Not much research has been done on perceptions of academic dishonesty policy, and the literature on cheating mostly includes self-report studies, where questions about students’ own cheating behavior are usually asked, and this can be unreliable. A study at a large midwestern public university with no honor code (Wajda-Johnston, Handal, Brawer, & Fabricatore, 2001) explored three aspects, among many, of dishonesty perceptions: prevalence of academically dishonest behaviors, seriousness and severity of the violations, and the reasons used to justify dishonesty. Based on that earlier research, the following hypotheses were made:
Hypothesis 1: It was hypothesized that, within an honor code environment, faculty would estimate a lower level prevalence of cheating than would students.
Hypothesis 2: It was hypothesized that, within an honor code environment, faculty would report the severity of honor code violations to be higher than would students.
Hypothesis 3: It was hypothesized that, within an honor code environment, students would endorse more justifiable reasons for cheating than would faculty.
Method
Participants
There were two types of participants in this quasi-experimental study. The first group was composed of Connecticut College students from introductory psychology courses (PSY 101 and PSY 102) who participated in this study for course credit. One hundred and twenty-nine students took the survey, but only 118 students were included in the analyses; 11 student responses were excluded because they failed to respond to 75% or more of the questions or excluded one or more sections of the survey completely. Of the students who opted to report their sex, 89 were women and 27 were men. Fiftyseven students were freshmen, 38 were sophomores, 17 were juniors, and 4 were seniors. The mean age of students was 19.34; ages ranged from 18 to 22.
The second group of participants was composed of Connecticut College faculty across all departments. With a response rate of 75.3%, 61 out of 81 faculty members contacted took the survey. However, 44 faculty members were included in the analyses. Seventeen faculty responses were excluded because they failed to respond to 75% or more of the questions or excluded one or more sections of the survey completely. Out of the faculty members who opted to report their sex, 24 were women and 15 were men.
Materials
Questions for this study were adapted from a study that measured academic dishonesty at the graduate level (Wajda-Johnston et al., 2001), which used a modified version of the Cheating/Academic Dishonesty Survey (LaGrange, 1992). There were two questionnaires, one for faculty (with 112 items) and one for the students (with 106 items). Questions were modified to better represent undergraduate students and the Connecticut College academic environment. No information on internal consistency was reported by the Wajda-Johnston et al (2001) study.
The student survey examined perceived prevalence, perceived severity, and perceived justification of honor code violations at Connecticut College. As mentioned before, these have been explored at a large non-honor university at the graduate level (Wajda-Johnston et al., 2001). Questions about students’ own cheating behavior were omitted from the original survey because of the lack of reliability these self-report measures possess in sensitive matters such as cheating. Examples of questions from the survey include “Getting high grades is more important than learning” (this was rated on a 5 point scale from “Not at all dishonest” to “Severely dishonest”) and “Suppose a faculty member saw or knew of a student engaging in academically dishonest behaviors. Choose the ONE action that you REALISTICALLY EXPECT the faculty member to do” (this was a multiple choice question and the options included: A. Immediately confront the cheater, B. Warn the class that he/she is aware that dishonest behavior is occurring in the class but not mention the cheater’s name, C. Threaten the cheater that you will report him/her to the dean or honor council if the cheating does not stop, D. Tell the cheater you saw him/her cheating and you want it to stop, E. Mention to the cheater you saw him/her cheating, F. Ignore the cheating and do nothing, G. Tell another faculty member that the behavior is occurring, H. Immediately report the cheater to the dean or honor council, and I. Notify the dean or honor council that the behavior is occurring but not mention the cheater’s name). Demographic information, such as age, sex, and class year was collected at the end of the survey.
Students first responded to prevalence items (items 1 to 37), followed by severity items (items 38 to 74), and then justification items (items 75 to 92). Items 93 to 96 asked general questions on what participants perceive the actions students and faculty would ideally and realistically take when aware of or confronted by a dishonesty case. Items 97 and 98 asked how concerned students and faculty are on campus about cheating. Items 99 and 100 asked how concerned students and faculty on campus should be regarding cheating. Items 101 to 104 are demographic questions. Item 105 asked students whether they have answered all questions honestly.
The faculty survey also examined the same three topics of prevalence, severity, and justification. An example of a question from the justification section of the survey includes “Because so many other students cheat, you could be hurting your own grades if you didn’t cheat too” (this was rated on a 5 point scale of “Not a valid justification” to “Very strong”). Another example of a question from this survey was “Have you ever been reasonably certain that a student in one of your classes at Connecticut College was aware that cheating was occurring but chose to do nothing about it?” This was a “Yes” or “No” question.
On the faculty survey, items 1 to 37 examined prevalence levels, items 38 to 74 examined severity, and 77 to 95 examined justification (analogous to the student survey). Items 75 and 76 examined how many times the faculty participants have witnessed cheating and what percentage of students do engage in academically dishonest behaviors. Items 96 to 99 included what students and faculty would realistically and ideally do in cases of
dishonesty. Items 100 and 101 explored whether faculty encountered situations where students and other faculty members have not taken any actions against cheating. Items 102 to 104 examined how concerned faculty and students are about dishonesty levels. Items 105 and 106 examined how concerned about cheating they should really be.
A reliability analysis was conducted on the modified scale, and it was found to be highly reliable overall, Cronbach’s α = .93. The separate subscales also were found to be reliable: prevalence subscale, Cronbach’s α = .97; severity subscale, Cronbach’s α = .97; justification subscale, Cronbach’s α = .95.
Procedure
Student signup sheets were posted on the second floor of Bill Hall at Connecticut College on a bulletin board next to other signup sheets for other ongoing research projects. Participants were PSY 101 and PSY 102 students. The survey was administered online through Qualtrics TM and the Connecticut College Survey Monkey TM account. URLs to the survey were emailed to each participant. The online survey required the students to fill out the informed consent form. A debriefing form appeared at the end of the survey, and participants were asked to print it out to obtain course credit for research participation.
Faculty surveys were sent out through email using a modified Dillman (2007) survey method. The first email directly has the URL to the questionnaire, and it requested participation and explained the importance of the study. Two subsequent weekly reminders were sent after the first email. Every survey had the informed consent form at the beginning and the debriefing form at the end.
Results
The purpose of this study was to examine the perceptions of prevalence and severity of academic dishonesty and of justifications given for academically dishonest behaviors across students and faculty at Connecticut College. To distill the ratings of prevalence, severity, and justification, the dimensions were identified through factor analysis, a process used to discover patterns of relationships among items, and create independently grouped subsets, which usually have an ascribed meaning and convey certain information about that particular subset. In this case, the grouping occurs based on the rating each cheating-related behavior received. The factor analysis method employed here involved a Principal Components Analysis extraction method and a Varimax with Kaiser Normalization rotation method, and this was conducted separately for each of the three components (prevalence, severity, and justification).
For prevalence, the initial solution produced five factors with eigenvalues greater than 1 (the rotation converged in eight iterations). Sixteen items did not meet the minimum loading criterion of .40, or loaded at that level or higher on more than one factor. The fourth and fifth factors were not well defined, with very few items per factor. Consequently, a subsequent analysis was done to extract and rotate three factors (the rotation converged in 5 iterations). The third factor was not used because it only contained two items, and it was not found to be prominent after examination of the scree plot. Fourteen of the original 37 items clustered on two factors that were labeled Category 1 (normal) and Category 2 (extreme), which accounted for 56.39% of the variance. The first factor, Category 1 (normal), accounted for 32.79% of the variance and contained 8 items. Normal dishonest behaviors grouped into this category. The second factor, Category 2 (extreme) accounted for 23.60% of the variance and contained 6 items. Extremely dishonest behaviors, including those involving monetary and sexual compensation, grouped into this category. Table 1 presents the items and factor loadings.
For severity, the initial solution produced three factors with eigenvalues greater than 1 (the rotation converged in four iterations). Twelve items did not load greater than .40 or loaded at that level or higher on more than one factor. The third factor was not very well defined, as it had one item. Consequently, a subsequent analysis was done to extract and rotate three factors again (the rotation converged in four iterations). The third factor was not used because there was just one item and it was not found to be prominent after examination of the scree plot. Twenty-three of the original 37 items clustered on two factors that were labeled Category 1 (extreme) and Category 2 (normal), which accounted for 62.32% of the variance. The first factor, Category 1 (extreme), accounted for 35.11% of the variance and contained 13 items. As with prevalence, extremely dishonest behaviors, including those involving monetary and sexual compensation, grouped into this category. The second factor, Category 2 (normal) accounted for 27.21% of the variance and contained 10 items. This factor contained “normal” academically dishonest behaviors. Table 2 presents the items and the factor loadings.
For justification, the initial analysis produced three factors with eigenvalues greater than 1 (the rotation converged in five iterations). Two items did not load greater than .40 or loaded at that level or higher on more than one factor. The third factor was not very well defined, as it had just one item. Consequently, a subsequent analysis was done to extract and rotate three factors again (the rotation converged in five iterations). The third factor was not used because there was just one item, which was not found to be prominent after examination of the scree plot. Sixteen of the 19 original items clustered on two factors that were labeled “grade/career-oriented” and “situational,” which accounted for 60.87% of variance. The first factor, grade/career-oriented accounted for 31.59% of the variance and contained 8 items. This factor contained items pertaining to student beliefs regarding the importance of good grades, regardless of how they are achieved. It also consisted of items that concern the importance of high grades for future career plans. The second factor, situational, accounted for 29.28% of the variance and also contained 8 items. This factor consisted of items concerning the education and examination environment and the punishments involved. Table 3 presents the items and the factor loadings.
Items that loaded on each factor were summed, divided by the number of items in each factor, and used in subsequent analyses. For each of the items in a given factor for prevalence, the higher the value, the more prevalent academic dishonesty is perceived to be (e.g., 1 = 0% and 11 = 91-100%). For each of the items in a given factor for severity, the higher the value, the more severe academic dishonesty is perceived to be (e.g., 1 = Not at all dishonest and 5 =Severely dishonest). For each of the items in a given factor for justification, the higher the value, the less justifiable the behavior is perceived to be (e.g., 1 = Not a valid justification to 5 = Very strong). Table 4 presents the overall means and standard deviations of the means of the factors in each factor (prevalence, severity, and justification).
To evaluate the first hypothesis that, within an honor code environment, faculty would estimate a lower prevalence of cheating than would students, a one-way multivariate analysis of variance (MANOVA) was conducted. The two levels of type (students and faculty) served as the fixed factor, and the means of the two factors obtained through factor analysis, Category 1 (normal) and Category 2 (extreme), served as the dependent variables. The analysis was significant, Wilks’s λ = .830, F(2, 156) = 10.64, p <.001. Univariate analyses of variance revealed significant differences for Category 1 (normal), F(1, 158) = 9.22, p = .003, but no significant differences were found for Category 2 (extreme), F(1, 158) = 2.71, p = .102, between students and faculty. In the normal category, faculty judged the prevalence of cheating to be less prevalent than did students. Means and standard deviations of the prevalence ratings, for both students and faculty, and the results of the univariate analyses are presented in Table 5.
To evaluate the next hypothesis that, within an honor code environment, faculty would report the severity of honor code violations to be greater than would students, a one-way MANOVA was conducted, with the two levels of type (students and faculty) as the fixed factor, and the means of the two factors obtained through factor analysis distillation, Category 1 (extreme) and Category 2 (normal), as the dependent variables. The analysis was significant, Wilks’s λ = .772, F(2, 159) = 23.52, p < .001. Univariate analyses of variance revealed significant differences for Category 1 (extreme), F(1, 160) = 7.14, p =.008, and Category 2 (normal), F(1, 160) = 46.16, p < .001, between students and faculty. In both categories, faculty rated the behaviors to be more severe than did students. Means and standard deviations of the severity ratings, for both students and faculty, and the results of the univariate analysis are presented in Table 6.
To evaluate the final hypothesis that, within an honor code environment, students would endorse more justifiable reasons for cheating than would faculty, a one-way MANOVA was conducted, with two levels of type (students and faculty) as the fixed factor, and the means of the two factors obtained through factor analysis distillation, grade/career-oriented and situational, as the dependent variables. The analysis was significant, Wilks’s λ =.670, F(2, 158) = 38.83, p < .001. Univariate analyses of variance revealed significant differences for grade/career-oriented, F(1, 159) = 68.67, p < .001, and situational, F(1, 159) = 14.10, p < .001, between students and faculty. Students provided higher ratings on the justification items (of both categories) than did faculty. Means and standard deviations of the justification ratings, for both students and faculty, and the results of the univariate analyses, are present in Table 7.
In addition, further analysis was done on the student data to examine whether there were differences between women and men. A one-way MANOVA with two levels of sex (women and men) as the fixed factor and the factors for justification as the dependent variables approached significance,Wilks’s λ = .950, F(2, 113) = 2.95, p = .056. An additional one-way MANOVA with two levels of sex (women and men) as the fixed factor and the factors for severity as the dependent variables also approached significance, Wilks’s λ =.952, F(2, 113) = 2.83, p = .063.
Certain questions that produced categorical data were also asked to both students and faculty near the end of the survey. One of these questions was “How concerned do the following groups at Connecticut College appear to be about cheating/academic dishonesty?” Among students, 42.7% thought that faculty care a great deal about academic dishonesty and 9.5% thought that students care a great deal about academic dishonesty; while among faculty, 41.5% thought that they themselves care a great deal about academic dishonesty, only 4.9% thought that students care a great deal about academic dishonesty, and 29.3% thought that other faculty care a great deal about academic dishonesty.
In addition, among students, 29.1% thought that faculty care a good deal about academic dishonesty and 33.6% thought that students care a good deal about academic dishonesty; while among faculty, 34.1% thought that they themselves care a good deal about academic dishonesty, 26.8% thought that students care a good deal about academic dishonesty, and 43.9% thought that other faculty care a good deal about academic dishonesty.
Moreover, among students, 17.1% thought that faculty somewhat care about academic dishonesty and 33.6% thought that students somewhat care about academic dishonesty; while among faculty, 19.5% thought that they themselves somewhat care about academic dishonesty, 46.3% thought that students somewhat care about academic dishonesty, and 17.1% thought that other faculty somewhat care about academic dishonesty.
Furthermore, among students, 8.5% thought that faculty minimally care about academic dishonesty and 18.1% thought that students minimally care about academic dishonesty; while among faculty, 2.4% thought that they themselves minimally care about academic dishonesty, 22% thought that students minimally care about academic dishonesty, and 7.3% thought that other faculty minimally care about academic dishonesty.
Finally, among students, 2.6% thought that faculty did not care about academic dishonesty at all and 5.2% thought that students did not care about academic dishonesty at all; while among faculty, 2.4% thought that they themselves did not care about academic dishonesty at all, no one thought that students did not care about academic dishonesty at all, and 2.4% thought that other faculty did not care about academic dishonesty at all.
Another question that produced categorical data was “Do you think the following groups at Connecticut College should be more concerned about cheating than they currently appear to be?” Among students, 6.8% thought that faculty should be much more concerned about academic dishonesty and 5.1% thought that students should be much more concerned about academic dishonesty; while among faculty, no one thought that they themselves should be much more concerned about academic dishonesty, 40% thought that students should be much more concerned about academic dishonesty, and 10.3% thought that other faculty should be much more concerned about academic dishonesty.
In addition, among students, 38.5% thought that faculty should be more concerned about academic dishonesty and 12% thought that students should be more concerned about academic dishonesty; while among faculty, 20% thought that they themselves should be more concerned about academic dishonesty, 42.5% thought that students should be more concerned about academic dishonesty, and 33.3% thought that other faculty should be more concerned about academic dishonesty.
Moreover, among students, 47.9% thought that faculty should not change their level of concern about academic dishonesty and 77.8% thought that students should not change their level of concern about academic dishonesty; while among faculty, 75% thought that they themselves should not change their level of concern about academic dishonesty, 15% thought that students should not change their level of concern about academic dishonesty, and 53.8% thought that other faculty should not change their level of concern about academic dishonesty.
Furthermore, among students, 3.4% thought that faculty should be less concerned about academic dishonesty and 2.6% thought that students should be less concerned about academic dishonesty; while among faculty, 5% thought that they themselves should be less concerned about academic dishonesty, 2.5% thought that students should be less concerned about academic dishonesty, and 2.6% thought that other faculty should be less concerned about academic dishonesty.
Finally, among students, 3.4% thought that faculty should be much less concerned about academic dishonesty and no one thought that students should be much less concerned about academic dishonesty; while among faculty, no one thought that they themselves, students, or other faculty should be much less concerned about academic dishonesty.
Over 85% of faculty members (n = 37) said that they address academic dishonesty on the first day of classes whereas 14% (n = 6) said that they do not. Over 80% of faculty members (n = 36) said that they address cheating in their course syllabi, whereas 16% (n = 7) said that they do not.
Discussion
Do students and faculty perceive academic infractions differently in the context of an honor code? Results indicate that compared to students, faculty members tend to estimate a lower level of prevalence of general academically dishonest behaviors. However, this was not the case with extreme cases of dishonesty: Faculty and students rated them similarly. Extreme cases of dishonesty were defined (through factor analysis) as those behaviors that included some form of monetary or sexual compensation, such as in cases where a student took an examination for another student. Both students and faculty did not perceive extreme cases to be very prevalent. These results of the present study are supported by those of Brimble and Stevenson-Clarke (2005). Wajda-Johnston et al. (2001) found significant differences on certain prevalence ratings between students and faculty, but concluded that they were not meaningful because, despite the significance difference, the means for both faculty and students were found to be in the same percentage range.
However, the current study revealed that the prevalence ratings, in addition to statistical significance, are meaningful differences. Students generally perceive prevalence of cheating to be in the 21 - 30% range, whereas faculty perceive it to be in the 11 - 20% range. A reason behind this difference may be that students are more likely than faculty to witness academic dishonesty first hand in situations such as unproctored examinations or take-home assignments. It also may be that faculty are not usually physically present in examination situations or dormitory and library environments, and that they place their full trust in students based on the honor code; consequently, they may not really be aware of much cheating and dishonesty. The ratings of prevalence found in this study are higher than those found in the previous study done at a graduate level in a midwestern, religiously affiliated university (Wajda-Johnston et al., 2001), which may suggest that dishonesty at the graduate level is perceived is perceived to be less prevalent that it is perceived to be at the undergraduate level. Moreover, students in the Wajda-Johnston et al. (2001) study may be more aware of the ethical/moral dimensions of cheating, because of the religious affiliation of the university.
Results also indicate that faculty reported honor code violations to be more severe than did students. This difference was present in both normal and extreme cases of dishonest behaviors. Extreme cases of dishonesty were rated more harshly in terms of severity than were normal cases of dishonesty. The mean ratings in both extreme and normal groups were very high across both groups of faculty and students. There were no ratings below 3.6 (on a scale of 5, where 5 is the most severe), and scores were rated “moderately dishonest” or “severely dishonest,” which shows that, despite their differences, both students and faculty in an honor code environment perceive academically dishonest behaviors to be quite severe and morally incorrect. However, the means (and the low standard deviations) did not seem to make a practically meaningful difference, because, after score conversion, both students and faculty in the extreme category rated the behaviors to be “severely dishonest,” and both students and faculty in the normal category rated the behaviors to be “moderately dishonest.” This result is supported by previous research (Wajda-Johnston et al., 2001). This result also seemed to be similar to another study that concerned severity perceptions (Sims, 1995), where it was found that faculty generally consider dishonest behaviors to be more severe than do students. A possible reason for faculty’s harsh judgment of academic dishonesty is because it is their duty to uphold the honor code, and they act as the prime administrators and implementers of the system. Also, faculty view academic work in a professional way.
For justification, students seem to endorse more reasons for cheating that did faculty. This was the case in both grade/career-oriented and Situational factors. Very few items were rated as “moderately strong” or “very strong” in terms of justification, which shows that, in general, both students and faculty are less willing to support reasons given as justifying academic dishonesty. The difference between faculty and students on grade/career-oriented ratings was found to be practically meaningful; faculty rated these items from “not a valid justification” to “very weak,” whereas students rated them from “very weak” to “slightly weak.” However, the difference between faculty and students on situational items was not found to be practically meaningful, and both students and faculty rated them from “not a valid justification” to “very weak,” which was slightly different from previous research (Wajda-Johnston et al., 2001) and showed that most of the items did not produce meaningful differences between students and faculty. The distilled justification factors produced are consistent with previous findings, which showed that competition for grades, insufficient study time, and desire for success are some reasons why students cheat (Baird, 1980; Stevens & Stevens, 1987). A possible reason why students endorse more academically dishonest behaviors as justifiable could be their stressful and competitive academic environment, which constantly compels them to strive for the best grades and academic opportunities. It could also be that they are more likely to have recently engaged in those behaviors, and therefore find them justifiable to relieve guilt.
Results from this study indicate that college administrators should reevaluate the honor code system as there is a clear discrepancy in how faculty and students perceive the honor code. It is alarming that students see academic dishonesty to be more justifiable and less severe than the faculty do.
Limitations and Practical implications
The current study has several limitations. The survey had certain questions that were aimed to examine cheating behavior in laboratory classes. However, such classes are mostly taught by faculty in the sciences. These questions, thus, were not applicable across all faculty. Moreover, not all professors at Connecticut College administer examinations or papers. Some choose one over the other, some administer both, and faculty in the arts assign other types of work. The faculty survey should have accounted for these particular cases. The number of participants in the faculty sample was not very high (N = 44). The response rate of faculty was fairly high, but it would have helped the power of the sample if more faculty had taken the survey. The student participants were skewed in terms of gender; 76.7% of the student sample consisted of women. Differences in perception of academic dishonesty between women and men in the student data were also examined, and the multivariate analysis approached significance (p = .056 for justification and p = .063 for severity). Significant results might have been revealed if there were a larger number of male participants. The student data also were skewed in terms of class year, as most introductory psychology students are freshmen and sophomores. Of the 116 students who chose to report their class year, 17 were juniors and only 4 were seniors. Being at the college and thus under the honor code for a longer time could affect beliefs about the system. The class distribution could be improved by conducting the survey campus-wide, instead of just restricting the participant pool to introductory psychology students.
It is also unclear whether the perceptions of prevalence given by students are reflective of their own cheating behaviors, as self-report of own behaviors was completely omitted from the study. Generally, self-report measures of cheating have revealed higher prevalence (McCabe, Butterfield, & Trevino, 2006).
Future Directions
The survey should be administered again to the entire campus community to achieve a wider participant pool. In addition to dealing with academic dishonesty, the honor code at Connecticut College deals with social and community infractions (Rodgers, 2001) and it could be valuable if violations of social aspects of the honor code were also examined. Comparisons could be made between perceptions of severity of academic honor code violation versus social honor code violations. In addition, borderline cases of academic dishonesty could be examined. These can be defined where there are no clear guidelines or rules in the honor code policy on whether that particular behavior constitutes as academic dishonesty or not. For this study, self-report measures of cheating were omitted because of the sensitivity of the issue. In future research, such self-report measures could actually be administered to look for differences between cheaters and non-cheaters in terms of their perceptions of prevalence, severity, and justification of academic dishonesty.
References
Baird, J. S. (1980). Current trends in college cheating. Psychology in the Schools, 17, 515-522. doi: 10.1002/1520-6807(198010)17:4<515::AIDPITS2310170417>3.0.CO;2-3
Brimble, M., & Stevenson-Clarke, P. (2005). Perceptions of the prevalence and seriousness of academic dishonesty in Australian universities. Australian Educational Researcher, 32, 19-44. doi:10.1007/BF03216825
Burns, S. R., Davis, S. F., Hoshino, J., & Miller, R. L. (1998). Academic dishonesty: A delineation of cross-cultural patterns. College Student Journal, 32, 590-596. doi:10.1080/10508420701713030
Connecticut College. (2012). Connecticut College student handbook. Retrieved from http://www.conncoll.edu/cw40/c/document_library/get_file?uuid=8eea2e03-4d6a-4d33-92bf-5f6abc8a3a47&groupId=10143
Del-Carlo, D. I., & Bodner, G. M. (2004). Students' perceptions of academic dishonesty in the chemistry classroom laboratory. Journal of Research in Science Teaching, 41, 47-64. doi:10.1002/tea.10124
Diekhoff, G. M., LaBeff, E. E., Clark, R. E., Williams, L. E., Francis, B., & Haines, V. J. (1996). College cheating: Ten years later. Research in Higher Education, 37, 487-502. doi:10.1007/BF01730111
Dillman, D. A. (2007). Mail and internet surveys: The tailored design method. Hoboken, New Jersey: Wiley Estep, H. M., & Olson, J. N. (2011). Parenting style, academic dishonesty, and infidelity in college students. College Student Journal, 45, 830-838. doi: 10.1007/s 10551-007-9576-0
Jordan, A. E. (2001). College student cheating: The role of motivation, perceived norms, attitudes, and knowledge of institutional policy. Ethics & Behavior, 11, 233-247. doi:10.1207/S15327019EB1103_3
Keith-Spiegel, P., Tabachnick, B. G., Whitley, B. E., & Washburn, J. (1998). Why professors ignore cheating: Opinions of a national sample of psychology instructors. Ethics & Behavior, 8, 215-227. doi:10.1207/s15327019eb0803_3
Kreit, B. (2001, September 21). Survey shows that friendships come before conscience. College Voice, p. 1.
LaGrange, M. V. (1992). Student, faculty, and administrator perceptions of academic dishonesty. Unpublished master’s thesis, Saint Louis University, Saint Louis, Missouri.
Malinowski, C. I., & Smith, C. P. (1985). Moral reasoning and moral conduct: An investigation prompted by Kohlberg's theory. Journal of Personality and Social Psychology, 49, 1016-1027. doi:10.1037%2F%2F0022-3514.49.4.1016
McCabe, D. L. (1993). Faculty responses to academic dishonesty: The influence Perceptions of Academic Dishonesty 70 of student honor codes. Research in Higher Education. 34, 647-658. doi:10.1007/BF00991924
McCabe, D. L., Butterfield, K. D., & Trevino, L. K. (2006). Academic dishonesty in graduate business programs: Prevalence, causes, and proposed action. Academy of Management Learning & Education, 5, 294-305. doi:10.5465/AMLE.2006.22697018
McCabe, D. L., & Trevino, L. K. (1993). Academic dishonesty: Honor codes and other contextual influences. Journal of Higher Education, 64, 522-538.
McCabe, D. L., & Trevino, L. K. (1996). What we know about cheating in college: Longitudinal trends and recent developments. Change, 28, 28–32.
McCabe, D. L., & Trevino, L. K., & Butterfield, K. D. (2001). Cheating in academic institutions: A decade of research. Ethics and Behavior, 11, 219-232. doi:10.1207/S15327019EB1103_2
McCabe, D. L. & Trevino, L. K., & Butterfield, K. D. (2002). Honor codes and other contextual influences on academic integrity: A replication and extension to modified honor code settings. Research in Higher Education, 43, 357-378. doi:10.1023/A:1014893102151
Never, B. (1999, February 26). Problems lie with the honor code, not confidentiality. College Voice, p. 2.
Preston, T., & Ammirati, J. (1991, September 3). Our school
and our honor code. College Voice, p. 2.
Pulvers, K., & Diekhoff, G. M. (1999). The relationship between academic honesty and college classroom environment. Research in Higher Education, 487-498.
Rodgers, J. (2001, November 9). J-board statistics reveal more male honor code violations. College Voice, p. 1.
Roig, M., & Marks, A. (2006). Attitudes toward cheating before and after the implementation of a modified honor code: A case study. Ethics & Behavior, 16, 163-171. doi:10.1207/s15327019eb1602_6
Simon, C. A., Carr, J. R., McCullough, S. M., Morgan, S. J., Oleson, T., & Ressel, M. (2004). Gender, student perceptions, institutional commitments, and academic dishonesty: Who reports in academic dishonesty cases? Assessment and Evaluation in Higher Education, 29, 75-90. doi:10.1080/026029303200015817
Sims, R. I. (1995). The severity of academic dishonesty: A comparison of faculty and student views. Psychology in the Schools, 32, 233-238. doi:10.1002/1520-6807(199507)32:3<233::AIDPITS2310320311>3.0.CO;2-H
Stephens, J. M., & Nicholson, H. (2008). Cases of incongruity: Exploring the divide between adolescents' beliefs and behavior related to academic dishonesty. Educational Studies, 34, 361-376. doi:10.1080/03055690802257127
Stevens, G. E., & Stevens, F. W. (1987). Ethical inclinations of tomorrow’s managers revisited: How and why students cheat. Journal of Education for Business, 63, 24-29. doi:10.1080/08832323.1987.10117269
Vandehey, M. A., Diekhoff, G. M., & LaBeff, E. E. (2007). College cheating: A twenty-year follow-up and the addition of an honor code. Journal of College Student Development, 48, 468-480. doi:10.1353/csd.2007.0043
Wade, P. (1987, November 3). The honor code analyzed: Is it only an ideal? College Voice, p. 4.
Wajda-Johnston, V. A., Handal, P. J., Brawer, P. A., & Fabricatore, A. N. (2001). Academic dishonesty at the graduate level. Ethics & Behavior, 11, 287-305. doi:10.1207/S15327019EB1103_7
Ward, C. (2000, May 8). Honor code review could spell the end for self scheduled exams. College Voice, p. 1.
Weaver, K. A., Davis, S. F., Look, C., Buzzanga, V. L., & Neal, L. (1991). Examining academic dishonesty policies. College Student Journal, 25, 302-305.
Whitley, B. E. (1998). Factors associated with cheating among college students: A review. Research in Higher Education, 39, 235-274. doi:10.1023/A:1018724900565
Woodsome, K. (1998, October 23). Honor code confuses Connecticut college students. College Voice, p. 1