Perceived seriousness of academic cheating behaviors among undergraduate students: an Ethiopian experience

The study was conducted to examine perceived seriousness of academic cheating behaviors among undergraduate students in an Ethiopian University. A total of 245 (146 males and 99 females) regular undergraduate students were randomly selected from three colleges: business and economics, natural and computational science, and social science found in a university. Data were collected using a survey. The results indicated that majority of the respondents rated most cheating behaviors as “serious” The study found that although students perceived the seriousness of most cheating behaviors, they continued to actively engage in cheating. Furthermore, significant differences were found in the perception of the seriousness of academic cheating behaviors amongst students according to field of study and gender. Therefore, as today’s undergraduate students are the employees of tomorrow and that the ethics they adopt and adhere to shape their behavior in the future, it is suggested that there is a need to make known to students the reality of academic integrity and to expose them to the consequences of violating students’ academic code of ethics.


Introduction
In many social and economic contexts individuals often face the choice to adopt different types of opportunistic or even illicit behavior to increase their welfare taking advantage of others for personal interests. It is common to see people cheating on taxes, free riding on public goods, claiming benefits without entitlement, bribing and corrupting public officials. Likewise, academic dishonesty is among these fraudulent behavior which is becoming a major concern and threat to education systems of almost all countries of the world.
Academic cheating as defined by Davis et al. (2009) is deceiving or depriving by trickery, defrauding, misleading or fooling another. They further explained that academic cheating/academic misconduct refers to acts committed by students that Academic cheating in higher education institutions is universally considered as a serious problem. However, most studies on the subject were undertaken in western countries (Yazici et al. 2011 andLim &See, 2001) and focused on the prevalence, determinants, and different definition of cheating (Jan et al. 2002;Fienberg 2009;and Witherspoon et al. 2012). Only a handful of studies have explored how people perceive cheating behaviors (Fienberg 2009). Students' perception regarding the severity of cheating behavior affects both the frequency and likelihood of the activity i.e. cheating behavior. For example, if a student believes that copying from friends on exam is trivial cheating, greater frequency of this activity would be reported. A student who does not perceive certain cheating behaviors as being unethical is more likely to cheat (Elias and Farag 2010).
It is obvious that differences in socio cultural settings, demographic composition and even educational policies and programs bring about difference in students' perception of cheating behavior. For example, cross cultural studies conducted to examine students' attitudes toward academic cheating have found evidence that students of different nationalities and of different cultures vary significantly in their perceptions of academic cheating (Godfrey et al. 1993;Jan et al. 2002;and Fienberg 2009).
Similarly, Diekhoff et al. (1999) found that Japanese college students, as compared to US students, report higher levels of cheating tendencies, have a greater propensity to neutralize the severity of cheating behavior and are not as disturbed when observing someone cheating in class. Moreover, a comparative study conducted by Lupton et al. (2000) found significantly different levels of cheating between Polish and US business students. The Polish students reported much higher frequencies of cheating than their American counterparts and were more likely to feel it was not so bad to cheat on one exam or tell someone in a later section about an exam.
Although these few studies conducted in other countries have shed some light on the cultural and national aspect of the issues, the researcher believed that there is a need for further studies within different contexts in order to widen the empirical base of cheating studies and improve our understanding of students' perception of cheating behavior. Furthermore, research examining cheating in general and students' perception of cheating in particular in Ethiopian higher education institutions is scant. Few studies conducted so far in Ethiopian higher education institution focused on prevention and detection (Mengistu, 2019;Wubalem et al. 2020), faculty perception (Tefera and Kinde, 2009), nature and causes (Mebratu 2016), and consequences (Nelson, Devardhi, and Dino;Solomon 2017). These studies have shown that academic dishonesty is a serious problem and widely spread among university students in Ethiopia with the prevalence rate of ranging from 53%-96%. In addition, the studies have suggested that much has to be done under the theme of academic dishonesty in Universities of Ethiopia. Therefore, this study was conducted to examine students' perception of cheating behavior in Dire Dawa University. The following research questions were raised.
How do university students perceive seriousness of academic cheating behaviors? Do students perceive seriousness of academic cheating behaviors differently according to their demographic background? What is the prevalence rate of academic dishonesty among Dire Dawa university undergraduate students?

Conceptual framework
Academic dishonesty is 'any fraudulent actions or attempts by a student to use unauthorized or unacceptable means in any academic work (Theart and Smit 2012). Jensen et al. (2001) also defines academic dishonesty as students' attempt to present others' academic work as their own. Academic cheating has two forms, which is cheating behavior such as copying answers of others and plagiarizing behavior such as citing without including the correct source. According to Craig and Dalton (2013), plagiarism includes intentional and unintentional actions in utilizing another person's work wrongly. It is conducted in the form of replicating another person's work, copying the whole text, or even buying another person's writing and then admitting it as one's own. It also refers an inaccurate and non-thorough behavior in quoting, citing, and reporting the source being used dishonestly (Spielberberger 2004).
On the other hand, cheating, as defined by Salkind (2008), a dishonest action with the element of deceiving with the goal of obtaining benefits or superiority from other students. Anderman et al. (2009) described cheating as four categories: information transfer between individuals, the use of assisting tools, exploitation of weakness, and copying answers or information. Cizek (2012) has defined academic cheating as any action taken before, during, or after the administration of a test and assignments that is intended to gain an unfair advantage or produce inaccurate results. Cizek (2012) further elaborated cheating as involving three domains: cheating by giving, taking, or receiving information from other during exam/test; cheating by using prohibited materials or information; and cheating to evade the assessment process. According to Davis et al. (2009) and Lambert et al. (2003), cheating includes, but is not limited to: lying; copying from another's test or examination; discussion at any time of questions or answers on an examination or test, unless such discussion is specifically authorized by the instructor; taking or receiving copies of an exam without the permission of the instructor; using or displaying notes, "cheat sheets," or other information devices inappropriate to the prescribed test conditions. For the purpose of this study, the author has used Anderman et al. (2009) andCizek (2012) definition and/or description of academic cheating as they are detail and best describe the nature of academic cheating common among Ethiopian students at all levels.
Academic dishonesty is not a new phenomenon, indeed it has been extensively studied in the academic context and empirical evidence indicates that it is changing and increasing in recent years (Davis et al. 2009;Yardley et al. 2009;Jensen et al. 2001;McCabe et al. 2001). Results regarding the influence of gender on cheating behavior have been mixed (Godfrey et al. 1993;Davis et al. 2009). Evidence for the gender differences in cheating vary with some studies showing strong differences and others showing no differences between males and females. A good number of studies (e.g. Kobayashi and Tedor 2012;Lupton and Chaqman 2002;Peled et al. 2013) have reported that academic cheating is more frequent in boys than in girls. On the other hand, Trost (2009) found no difference between male and female. Some other study by Jacobson and his colleagues as cited in Jensen et al. (2001), in contrary, reported that females are more likely to cheat than males.
Furthermore, other studies have demonstrated that females tend to rate most academic cheating behaviors as "most serious" and unethical than males (Fienberg 2009;Whitley 2001;and Whitley et al. 1999). However, despite their negative attitudes towards cheating behaviors, females were found to be engaged in cheating behavior just as frequently as males. A Meta analysis conducted by Whitley et al. (1999) and Fienberg (2009) reported that although women show a more ethical attitude toward cheating than men, "in practice, women were almost as likely to cheat as men". Wilkinson (2009) suggested that a significant element in the issue of academic cheating is the students' beliefs about right and wrong, and the level of seriousness attached to specific instances of wrong-doing. Moreover, McCabe and Trevino (1993) explained that as the perceived severity of punishments increases, the levels of individual cheating will become lower.
On the other hand, Elias and Farag (2010) contend that business students with a love of money do not perceive certain cheating behaviors as being unethical, and are therefore, more likely to cheat. Morris and Kilian (2006) confirm this argument with their study conducted in seven US universities. They have found that undergraduate accounting majors and business major students report higher frequencies of cheating than other major students. Therefore, Oneill and Pfeiffer (2012) concludes that if accounting students lack academic integrity, then accounting scandals are not surprising. They suggest appropriate training in ethics is necessary for business students.
Researchers have shown that dishonesty in college, cheating in particular, is a predictor of unethical behavior in subsequent professional settings Chaqman 2002 &Lim andSee 2001). In other words, people who engaged in dishonest behaviors during their college days continue to do so in their professional careers. Lupton and Chaqman reported that successful cheating behaviors in college carry over as a way of life after college. The works of Sims (1993) also confirmed this fact. Sim has found that people who engaged in dishonest behaviors during college are more likely to commit dishonest acts of greater severity at work.
The literature on students' perception of academic dishonesty include aspects such as what constitute cheating, seriousness of cheating behavior, why and how cheating occurs, and how cheating can be discouraged. This study, however, is delimited to the study of undergraduate students' perception about the seriousness of academic cheating behaviors among undergraduate students. Moreover, of the three types of academic dishonesty, i.e. exam-related, assignment-related, and research-related, usually mentioned in different literatures, exam and assignment related academic dishonesty were considered in this study.

Methods and materials
This study used descriptive survey research method. A total of 245 (146 males and 99 females) senior undergraduate students were selected from three colleges (Social Science and Humanities, Business and Economics, and Natural and Computational Sciences) found within a University at Dire Dawa City, Ethiopia, using stratified and simple random sampling. In the selection of participants, they were told that their participation is voluntary and the result will be reported in aggregate. To measure students' perception about the seriousness of academic cheating behavior, the researcher adopted academic dishonesty questionnaire used by Oneill and Pfeiffer (2012) and Witherspoon et al. (2012). Eleven items that fit to the context of Ethiopian Universities were selected from the instruments. In other words, academic cheating behaviors that are common and in Ethiopian universities are considered. For instance, cheating using internet is currently impossible because of no access to internet in examination rooms. Few of the items were modified based on the review of literature and experience of the researcher as University instructor. All the items are considered as academic cheating.
In the questionnaire, students were asked to indicate how often they had undertaken a particular cheating behavior (rated in three scale i.e. 0 = never, 1 = once, 2 more than once) and the degree to which they perceived the seriousness of the cheating behavior (rated in five scale i.e. 0 = not at all cheating, 1 = trivial cheating, 2 = moderate cheating, 3 = serious cheating and 4 = most serious cheating). The lower the value of the perception shows that the students accept the behavior as normal (acceptable).
Pilot testing was made to test and fit the instrument with the context of Universities in Ethiopia. Forty (40) students from the institute of technology participated in the pilot test and the Cronbach alpha coefficients of all the items included in the main study were acceptable with the coefficient values of 0.80 and above. Finally, as gaining accurate data on cheating is difficult, the researcher went to some effort to make sure that the students answered frankly and honestly. The students were informed at the beginning of the data collection sessions about purpose of the study and anonymity of their responses. In the actual data collection, the questionnaires were administered while students were in classroom, before the beginning of new session.

Results
The purpose of the study was to assess perceived seriousness of academic cheating behaviors among undergraduate students. Depending on the nature of the data, quantitative data analysis techniques such as frequency, percentage, mean, independent t-test, one-way ANOVA, and chi-square were employed. Table 1 presents the perceived seriousness of academic cheating behaviors as rated by the respondents. Table 1 summarizes the mean perceived seriousness ratings for items pertaining to respondents' attitudes toward academic cheating behaviors. The findings indicate that respondents surveyed generally perceived cheating behaviors to be "serious" (with a mean of 2.87). Not all cheating behaviors were perceived to be equally serious. The findings show that allowing someone to copy one's answer during a test or examination; doing other students' course work/assignment; and whispering the answers to friends during examination had mean scores below 2 (M = 1.69, M = 1.88, and M = 1.73 respectively) suggesting that these activities were considered only as moderate cheating by the respondents. Whereas cheating behaviors such as failure to contribute one's fair share in group project (M = 3.46) and allowing own course work/assignment to be copied by another students (M = 3.46) were perceived as most serious academic misconduct.
A one way analysis of variance (ANOVA) was conducted to compare mean score of students among the three colleges: Social Science, Business and Economics, and Natural and Computational Sciences. As can be understood from Table 3, there were statistically significant difference between the mean score of students' perceived seriousness of academic cheating behaviors among the colleges (2, 242) = 7.754, p = 001: the result of the tukey HSD indicate a statistically significant difference between college of Business and Economics and the other two colleges. Table 4 above summarizes the percentages of respondents who admitted to having engaged on at least one occasion in each cheating activity surveyed. The majority of respondents (81.26%) indicated that they had committed one of the surveyed cheating activities at least once; only 18.74% reported that they had never been involved in any form of cheating. This is an indication that cheating behavior was highly prevalent among the students participated in the study. This finding was congruent with the past study carried out in Addis Ababa and Jimma Universities by Tefera and Kinde (2009) who found that 82-96% of the students had committed one form of cheating at least once.
Descriptive statistics depicted in Table 4 suggest that a high percentage of respondents (about 93%) reported having allowed others to copy their answers during test or examination. Approximately 93% admitted that they have done an assignment/coursework for other students. Again more than 91% of the respondent reported that they have whispered answers to their friends during examination at least once. These findings might not be surprising as these behaviors rated as moderate cheating activities by the students (Table 1). Using text message to get/send answers in examinations recorded the lowest frequency at 16.33%. The study also examined whether there is an association between variables like gender and field of study (stream) and cheating behavior. To do this, a chi square test was applied and the results are indicated in the following tables Table 5.
Regarding gender difference in prevalence of cheating, the study revealed no difference between males and females. Approximately 81% of male respondents reported having cheated at least once. Similarly 80% of female respondents admitted that they have engaged in one of the cheating behavior at least once. The chi-square test is found to be not significant at x 2 = 0.222, df = 1, p > 0.05. This shows that although females perceived academic cheating behaviors more serious than their counterpart, the engaged in those activities as equal as males Table 6.
As can be seen from Table 6 above, the results of the chi square test demonstrated that there is a significant difference among colleges in self-reported cheating behavior. The chi square test among the three colleges were statistically significant at x 2 = 6.530, df = 2, p < 0.05. To examine which college's students involve more in cheating behavior, further analysis of standardized residual test was computed. The result is indicated in Table 7 below. As shown in Table 7, there is a significant difference between business and economics students and the remaining two colleges in cheating. This suggests that students of business and economics have engaged more in cheating behaviors than students of college of natural & computational sciences and social science & humanities. In other words, business and economics students engaged more in cheating behaviors than college of natural & computational sciences and social science & humanities.

Discussion
The study revealed that majority of the respondents perceived most cheating behaviors as serious problems. Respondents perceived cheating behaviors such as doing less than one' fair share in group assignment; using text messages to send/ receive answers; taking crib notes to the exam halls; and allowing own coursework/ assignment to be copied by others to be more serious. Such finding is congruent with previous researches (Yazici et al. 2011;Yardley et al. 2009;McCabe et al. 2001;Lim and See 2001;and Peled et al. 2013).
It is interesting that "cheating in the form of allowing someone to copy during exam"; "doing homework for others"; and "whispering the answers to friends during examination" were perceived by respondents to be least/moderate cheating and accordingly more incidence of these behaviors were reported. It is clear that cheating would be more likely to occur where students view cheating as a minor problem. This finding is congruent with the study of Oneill and Pfeiffer (2012) who found that allowing  someone to copy one's answers during a test or examination and whispering the answers to friends during an examination are considered as trivial cheating, accordingly, greater frequency of this activity was reported. Similarly, Graham (1994) reported that students with lenient attitudes towards academic cheating behaviors reported more cheating as compared to students with stricter attitudes. The reason why this happen might be beyond the scope of this study, however, one possible explanation according to McCabe (2005) is the probability of being caught and severity of the punishment for that specific behavior. Trost (2009), for example, contend that students believe that some academic cheating behaviors are more acceptable or tolerated than others because of the variations in disciplinary consequences of the academic dishonesty. In addition, Hutton (2006) found that students believed that engagement in academic cheating behaviors in school were alright because they saw the school doing nothing towards the behavior. Similarly, Oneill and Pfeiffer (2012) underlined that the probability of being caught and penalized are the most effective predictors of student engagement in academic cheating. The severity of punishment for cheating is the value which a student weighs the benefits he/she will gain from not being caught. To put it differently, some students may rationally assess the costs and benefits of their actions. However, despite their perception, still majority (81%) of the respondents reported that they have engaged in cheating behavior at least once. In other words, though the students perceive cheating as serious problem and unethical, they couldn't be deterred from cheating. This finding is in agreement with what Tefera and Kinde (2009) have reported. Cheating behaviors such as copying from a nearby student during a quiz, test, or exam; allowing someone to copy your answers during a test or examination; writing on the hands, chairs, walls of exam halls; doing other students' coursework/assignment for them; and whispering the answers to friends during an examination are the top five cheating behaviors with high rate of incidence. Wilkinson (2009) and Lupton and Chaqman (2002) also reported that copying from another student in an exam and/or assignment are forms of misconduct that have become areas of increasing concern for academics in higher education.
According to Wilkinson (2009) andMcCabe et al. (2001), though majority of the students indicated academic cheating behaviors as unacceptable behavior, they do not condone cheating. Davis et al. (1992) also reported inconsistency between students' belief about the seriousness of academic cheating behaviors and their actual behavior. Davis and his colleagues found that around 90% of students affirmed that cheating is incorrect but 76% of them admitted they have cheated at least once. In this regard, as discussed above, the likelihood of getting caught, instructors' supervision, and the severity of the punishment matters a lot in students' behavior.
In addition, the study compared the ratings of female and male students to determine if there are gender differences in beliefs about seriousness of cheating behaviors. Accordingly, female students consistently rate the items as more serious than male students. In other words, females have shown more ethical attitudes towards most academic cheating behaviors than males. "Doing less than one's fair share of work on a group project; copying from a nearby student during a quiz, test, or exam; using text messaging to get/send answers on test; working on an assignment in group when individual work is assigned; and copying homework from another student are the activities where females showed more ethical attitude than their male counterpart. A metaanalysis conducted by Whitley et al. (1999) and Fienberg (2009) also affirmed this finding. The irony here is that however, females have perceived most cheating behavior as serious and inappropriate acts, the study revealed that they got engaged in cheating as frequently as males. This finding is against the findings of most previous researches who found that males are more likely to cheat than females (Kobayashi and Tedor 2012;Oneill and Pfeiffer 2012;and Peled et al. 2013). On the other hand, the finding is congruent with previous studies (e.g. Anitsal et al. 2009;Monahan et al. 2018;Quaraishi and Aziz 2017;Whitley et al. 1999;Fienberg 2009) who have reported no difference between males and females in cheating. Such inconsistency, thus, strengthen Klein et al.'s (2007) argument that explain gender as an inconsistent determinant of academic dishonesty.
The study also revealed that business and economics students have shown less ethical attitude towards academic dishonesty than students from the school of social science and humanities and school of natural and computational sciences. Moreover, the students of business and economics students have engaged in cheating more than students in the school of social science and humanities and natural and computational sciences. The same was found by Elias and Farag (2010). According to Elias and Farag, business students, with love of money, do not perceive certain cheating behaviors as unethical, and therefore, more likely to cheat than others. Another study by Morris and Kilian (2006) also found that undergraduate Accounting majors and business major students report higher frequencies of cheating than other major students. A recent study by Dyer et al. (2020) have also concluded that academic honesty runs particularly high among business students.

Conclusion and implications
This study presented some limitations. First, data are based on self-reports which would may result in the vulnerability to socially response bias because of the sensitive nature of the topic. This may have resulted in under-reporting of cheating during tests and examinations. The researcher attempted to avoid this problem by guaranteeing complete anonymity and emphasizing the importance of honest answers to the questions. Future investigations shall collect such data also from other sources (e.g., institutional reports related to honor code violations) and informants (e.g., course mates and teachers). Second, being conducted in a single University and the small sample may not be representative of University students in Ethiopia. Thus, findings of this study should be replicated on other Universities with large number of students and varied compositions for comprehensive understanding of academic cheating. Despite the limitation, however, this study has shed some light with regard to how university students perceived academic cheating behaviors in the Ethiopian context.
From the findings, it would be concluded that cheating is rampant in the university and students seems to be tolerant of unethical academic behaviors. This conveys a clear message that academic institutions has revisit their academic integrity policies and take necessary measures to deter academic cheating as its impact goes beyond the individual impact of crossing a moral or ethical boundary. Chace (2012) and Mensah et al. (2016) underlined that it reduces the perceived academic integrity of the institution, devaluing degrees earned from that institution. Similalry, Wollack and Cizek (2017) added that academic cheating threatens the validity of those credentials. Students who cheat rather than learn to pass courses are less prepared for the workforce and are more likely to engage in behaviors that are similarly unethical (Smyth et al. 2009;Teixeira and Rocha 2010). In view of this, the researcher discussed the implications as follows; The researcher believes that high prevalence of the cheating as witnessed in the study might be attributed to the less probability of getting caught and severity or absence of punishment. Thus, the university has to revisit the strength and applications of its students' code of conduct. As argued by Vandehey et al. (2007), students' concern for being caught and punished for cheating was far more effective as a deterrent to academic misconduct than the student's internal feelings toward the act. Vandehey et al. (2007) further emphasized that student punishment factors continued to be the best deterrents to student academic misconduct. If students are caught cheating, accused, and have a severe admonishment, it is more likely that, according to Petress (2003), the behavior can be curtailed. Levy and Rakovski (2007) also added that students' engagement in academic dishonesty decreased when they knew the cheating would be discovered and severely punished.
Moreover, given the essential nature of academic integrity to the academic mission of an institution, preventing academic dishonesty on the most common form of assessment (testing) is of high value to many colleges and universities. In classrooms and in the test center environment, this threat to academic integrity should lead to very strict security rules. Students should be observed at all times while taking test and/or examination and invigilators must be able to intervene immediately if there is any unusual testing behavior.
Above all, following an educational approach would be noteworthy to bring about long-term and sustainable changes within the minds of our students. Accordingly, the integration of academic integrity into every curriculum would seem to provide a potential means of increasing students' ethical sensitivity and, hence, behavior. Emphasizing this, Nonis and Swift (2001) have stated that we can implement fundamental approach involving the teaching of integrity and discussion of ethical issues in every subject, with particular emphasis in the capstone subjects instead of integrating a stand-alone ethics education/subject. Such an approach, according to Nonis and Swift (2001), focuses not so much on teaching students the rules of ethical analysis, but on providing a strong ethical foundation that becomes deep-rooted within students' personality over the duration of their studies so that they come to value integrity and honesty in their own behavior. The last, but not the least, more researches on academic integrity at all levels of the education ladder shall be conducted so that timely and preventive alternatives shall be considered. Moreover, higher education institutions and/or other stakeholders of the education sector shall encourage such studies through funding and dissemination thereby maximizing quality education.