An investigation on Turkish EFL Majoring Students ’ Unethical Behaviors

Much research has investigated plagiarism in different educational contexts; however, very little research has investigated English-as-a-foreign language students’ ethic from gender and impression management perspectives in Turkish EFL context. Academic dishonesty scale, impression management scale, and personal attributes questionnaire were used to collect data. This survey investigates Turkish EFL-majoring students’ attitudes towards academic behaviors, impression management, and psychological gender. Data were collected from 285 EFL majoring students from 3 state universities in Turkey. The results of this present study showed that first, all students have a moderate level of cheating behavior, they have varying level of beliefs about cheating behavior, there are no significant cheating behavior differences by gender. Anahtar Kelimeler etik davranışlar YDİ öğrenenler intihal üniversite öğrencileri akademik sahtekârlık


Introduction
"Plagiarism, collusion and other forms of academic misconduct have always been a regrettable but unavoidable aspect of academic existence" (Carroll & Appleton, 2001, p.6).However, plagiarism is not an easy concept to identify when considering various profoundly contextualized authorship and textual practices in a variety of domains and organizations (Hodges, Bickham, Schmidt & Seawright, 2017).Since textual regulations of approved submissions have changed over time and they show varies from context to context, even within the same social community, it is getting more problematic to define plagiarism (Hodges et al., 2017).Coughlin (2015) states that defining the concept of plagiarism in a clear way is not the core of the problem, but it is constructing concurrence and having learners and lecturers adopt the definition and implement necessary disciplinary and educational steps with a sensible and sustainable manner.
In the Compact Oxford English Dictionary (2009) the concept of plagiarism is identified as 'the practice of taking the work or idea of someone else and pass it off as one's own'.Angelil-Carter (2000) asserts that the word of plagiarism has meanings of 'kidnap' and 'plunder' in the Latin and it has some criminal connotations (p.16).In the literature, the concept of plagiarism is defined as literary theft (Park, 2003), academic dishonesty (Hollinger & Lanza-Kaduce, 1996) or academic misconduct (Stern & Havlicek, 1986).Pavela (1978) states four types of academic misconduct: cheating (using unauthorized material intentionally), fabrication (unauthorized falsification) or invention of any information or citation, facilitation (intentional attempt to help another by violating the rules) and lastly 'the deliberate use, adoption or reproduction of ideas, words or statements of another person as one's own without acknowledgement of the author' (p.73).Alzahrani, Salim and Abraham (2012, p.134) address two types of plagiarism: literal plagiarism and intelligent plagiarism.The former refers to the simple act of copying a text available on the internet and pasting it without doing any changes and using quotations) while the latter refers to a serious intentional act of deceiving readers by manipulating the original material without citing the author.On the other hand, there are three types of plagiarists: the first one is 'the accidental plagiarist' who does not know appropriate rules governing textual borrowing.Second type of plagiarist is 'the opportunistic plagiarist' who is aware of the issues of plagiarism but engages in misconduct due to fear, ethical lapses, disorganization, and others.Finally, the last type plagiarist is 'the committed plagiarist' who purposefully cheat and steal from others (Beasley (2004;p. 9).Jolly (1998) points out that there might be various underlying reasons for academic dishonesty such as an intentional act of deceiving and unintentional act of not implementing the rules of academic writing appropriately.In addition, Park (2003) restates that while intentional plagiarism is the student's purposeful act of deceiving by stealing from other sources, unintentional plagiarism is the student's failure of implementing appropriate textual citation protocols.Thomas (2000) suggests that increasing availability of mass-produced books and materials increased the rate of intentional plagiarism or stealing accordingly."Social media and collaborative websites also allow students to access, store and share class materials (e.g., previous year's exams/assignments, assignment materials, etc.)" (Sayed & Lento, 2015, p.74).Moreover, some of the recently developed websites offer the opportunity of purchasing readily prepared homework papers and solution manuals for course books (Sayed & Lento, 2015).Thus, the need for processing suspected large documents aroused the development of computerized methods (software programs such as Turnitin, Ithenticate, etc.) for detecting plagiarism.In this context, it can be said that not only students facilitate from the opportunities of technology and get involved in plagiarism, but also professors make use of technological tools to detect and deal with academic plagiarism (Sayed & Lento, 2015).
"Plagiarism by students is a moral maze because it raises important ethical and moral questions about good/bad or right/wrong behavior and about acceptable/unacceptable practices."(Park, 2003, p.474).According to Carroll and Appleton, in most cases, 'plagiarism' includes every action that a university student is not allowed to do during his/her higher education (2001, p.13).Students generally plagiarize in certain ways such as paraphrasing texts by not acknowledging the original author, submitting others' works as their own, copying the same material but showing it as a paraphrased form and stealing from other sources and using it as their own (Park, 2003;Wilhoit, 1994).Some students are more prone to cheat than the others are and among the international students who deal with language difficulties constitute a risk group, as English is not their native language (Park, 2003).Hodges et al. (2017)indicate that many non-Western nations do not inform students clearly about the ethical standards of publication and as their cultural perspectives and approaches towards plagiarism or patchwriting are different from the ones of western nations, they are 'more likely to be caught by Turnitin and other text matching software' (p.2).In this context, O'Neill (2012) claims that most of the students who were caught when attempting or committing plagiarism or misconduct declared that they were unaware of the seriousness of their actions and they have various understandings regarding academic dishonesty.According to Pennycook (1996), although many non-western students do seem to be aware of problems with plagiarism, they mostly do not know specific procedures of textual borrowing and how to prevent misconduct.
Contextual variables such as size and physical conditions of classrooms and personal characteristic and applications of instructors were previously determined to have an impact on academic misconduct (McCabe & Trevino, 1993).Over the past few decades, a number of studies have examined the academic dishonesty of college and university students and some of them have shown that not only contextual factors but also characteristics of the individuals have a significant impact on their tendency to apply academic misconduct (Fezatte 2009;Eriksson & McGee, 2015;Lambert & Hogan, 2004;McCabe, Trevino & Butterfield, 2001).Individual differences of students such as self-efficacy, having an instrumental or integrative motivation, personal features and attitudes have been involved in factors that contribute to a student's motive for practicing academically unethical behaviors (Fezatte, 2009;Jordan, 2001;Murdock, Hale & Weber, 2001).Whitley (1988) asserts that the attitudes of the students towards unethical behaviors might provide us with an insight into the predictable severity of academic dishonesty.For example, Bolin (2004) investigated whether there is an association between self-control, attitudes and academic misconduct and the researcher determined that in this trio attitudes had a central role.
A large body of studies has investigated the level of cheating in higher education institutions and it seems that the severe of academic misconduct shows variety from study to study (Becker & Ulstad, 2007;Bower, 1964;Macaulay & Lemn, 2014;McCabe & Trevino, 1993;Stern & Havlicek, 1986).For instance, a study conducted by Bower (1964), 75% of the participating university students had admitted that they cheated at least one time during their college education.In another study carried out by Macaulay and Lemn (2014), 80% of the students accepted that they engaged in some sort of academic misconduct.Again, in a study conducted previously, it was found that 80% of the students had cheated during their college careers (McCabe & Trevino, 1993).Stern and Havlicek (1986) reported that more than two-thirds of the students participated in their study admitted to applying some sort of cheating behavior during their higher education careers.In a similar vein, Becker and Ulstad (2007) examined 515 university students in order to understand the manners they adopt towards academic dishonesty.The results of the study revealed that most of the students who participated in the study accepted that they applied various cheating practices.However, most of these students were not able to identify some of the unethical behaviors correctly.
Plagiarism and other practices of academic misconduct constitute a growing problem as these actions spoil the coherence of educational process and probably debilitate the reliability of educational programs (Witmer & Johansson, 2015).At this point, a significant number of studies have provided considerable advice on the ways dealing with plagiarism regarding educative, preventative and punitive precautions (Beasley, 2004;Bretag & Mahmud, 2009;Carroll & Appleton, 2001;Coughlin, 2015;Park, 2003;Sayed & Lento, 2015).The recommendations and strategies that previously held studies have suggested for coping with plagiarism were not only at the instrumental level but also at the level of individual educator, and most of these suggestions coincide with each other (Bretag & Mahmud, 2009).For instance, the improvement of electronic software tools has provided a practical way of determining and preventing student plagiarism (Bretag & Mahmud, 2009).Bretag and Mahmud (2009) assert that instructors might prevent student plagiarism by using electronic software tools within an educated framework by enabling students to take over the responsibility of their works and by keeping up with their progress during the stages of drafting.According to Carroll and Appleton (2001), the most efficient way of coping with plagiarism is integrating academic and policy resolutions in a neutral, judicious and compatible way and constructing such a balanced attitude should incorporate the procedures of redesigning the course, providing students with adequate information about adjustments and having them acquire appropriate citation and academic discourse skills.Sayed and Lento (2015) suggest that instructors can cope with plagiarism and cheating by not using the same assessment materials repeatedly and by creating different exam papers, tasks, and assignments each year.Further, there is still non-consensus whether females cheat less than males.Therefore, this paper contributes to the literature by asking this guiding research question: What are the attitudes of Turkish EFL male and female students towards unethical behaviors?

Participants
Two-hundred-eighty-five undergraduate English-as-a-foreign language students studying English language teaching and English language and literature departments in four Turkish state universities participated in this study.Participating was voluntary.The following Table 1 and Table 2 show the details about the profiles of the participants.Again, Table 2 shows the participants class year by universities.As this study used convenience-sampling strategy, the numbers of the participants are unbalanced.First-year students are not included in this study because academic research courses are not given that year.

Data Collection Instruments
The following data collection instruments were taken from Backer and Ulstad ( 2007)'s study: The Academic Dishonesty Scale (ADS) (Spence et. al., 1975): It includes 9 Likert-type academically dishonest behaviors items.Psychological genders regarding instrumental expressiveness strengths are measured by the PAQ (Table1).In a study by Backer and Ulstad (2007), it was reported that the Cronbach alpha value is .90.Each student's ratings of the nine items were summed to produce one variable (CHEAT).The higher the value of CHEAT, the more accepting the student was of the cheating behaviors.
Impression management: It aims to measure whether a student is engaging in impression management.The twenty items are from the Balanced Inventory of Desirable Responding (BIDR), version 7 (Paulus, 1998cited in Backer & Ulstad, 2007).This scale aims to measure the differentiating fakers from non-fakers; that is, it helps to determine if the participants are deliberately heightening their responses when filling questionnaires (Paulus, 1998cited in Backer & Ulstad, 2007).The Cronbach Alpha value is .83.If the odd-numbered items are answered as 1 or 2, a student attempts to make a good impression.On the other hand, if the even-numbered items are answered as 6 or 7, a respondent attempt to make a good impression.The overall sum produces one IMR for each participant, ranging from 0-20.If a participant receives higher the IMR, the person engages more in impression management.
Personal Attributes Questionnaire (PAQ) (Spence et. al., 1975cited in Backer & Ulstad, 2007): It measures students' instrumental and expressive traits (see Figure 3).Backer and Ulstad (2007) reported the scale reliability .76(Cronbach Alpha of .76).Instrumental traits and expressive traits can be measured.The eight instrumental items are related to INSTRUM variable and the eight expressive items are related to EXPRESS variable.For each scale, a student's total responses to items are used to create one rating ranging from 8 to 40.

Data Collection Procedure
Two-hundred-eighty-five students completed the questionnaires during class time.They were voluntary and anonymous.First, the students' demographic information was obtained.Next, their attitudes toward academic behaviors, impression management, and psychological gender were measured.

Data Analysis
Descriptive statistics were used to analyze mean scores and standard deviations regarding the scale items and gender variables.Inferential statistics (e.g.t-test and ANCOVA) were used to examine significant differences between males and females in terms of their academic dishonesty levels, impression management and cheat behaviors.Table 3 shows the descriptive statistics about the CHEAT variable regarding the students' beliefs about cheating.Each student's total cheating rating was computed as one score by summing the student's ratings of the nine items.The mean scores are nearly 2 points, indicating that all students have a moderate level of cheating behavior.Further, the standard deviations are over 1 point, indicating that the students have a different level of beliefs about cheating behavior.Table 4 shows descriptive statistics regarding beliefs about cheating by gender.The results show that the mean scores and standard deviations across all items are very similar indicating that male and female students have a very similar level of CHEAT behaviors.Table 5 shows the t-test results regarding the students' CHEAT behaviors by gender.The results indicate that there is no significant difference between male and female students in terms of cheating behaviors (p=>.05).To investigate the relationship between students' gender and their ratings of cheating behaviors, the data was obtained from PAQ.In the PAQ, the strength of desirable, socially oriented traits which are attributed to female characteristics is measured by the expressiveness scale.However, the strength of expressiveness conditioning mostly determines ethical attitudes, indicating that if the students have a higher level of expressiveness, they probably less accept cheating behavior (Backer & Ulstad, 2007).Table 6 shows the descriptive analysis of the expressiveness (EXPRESS) of the students.EXPRESS was obtained by summing each student's responses the expressiveness questions in the PAQ.The mean scores for males and females are near to 4 score point, indicating that participants have a higher level of expressiveness rating (e.g. with a higher level of female ratings).Further, the standard deviations are lower than 1 score point, indicating that the participants have a similar level of expressiveness.More detailed analyses were performed to examine significant differences regarding expressiveness and gender relation.The covariate analysis of the expressiveness (EXPRESS) of the students was presented in Table 7.The Table 7 shows that gender is not significant (p>.01 but EXPRESS is a significant covariate in the analysis (p=.001), indicating that there is no gender effect after the inclusion of the student expressiveness in the analysis.

Table 3. Descriptive statistics for the participants' ratings in the Academic Dishonesty Scale (CHEAT variable)
To investigate the students' gender difference in the ratings of cheating behavior if the impression management is considered, a covariate analysis was performed.The results are presented in Table 8 and Table 9.Table 8 shows mean scores and standard deviations for IMR by gender.The females have lower mean scores than the male have.The standard deviations are over 1 score point, indicating that they have a different level of impression management.Table 9 shows the covariate analysis the gender effect and impression manage efforts of the participants.The results indicate that gender and impression management do not have any significant effect on cheating behavior.

Discussion and Conclusions
This study investigated if Turkish EFL male and female students have a different cheating rating and if they were engaging in impression management with their survey answers.To answer these questions, three scales were used (e.g.ADS, PAQ and IMR).The results of this present study showed that first, all students have a moderate level of cheating behavior and they have varying level of beliefs about cheating behavior.This result is supported by a study conducted by Bower (1964) in which 75% of the participating university students had admitted that they cheated at least one time during their college education.Further, there was no significant difference between male and female students in terms of cheating behaviors (p=>.05).Regarding the unethical behaviors, males take more risk in unethical behaviors than females (Becker & Ulstad, 2007;Webber, Blais & Betz, 2002).
Next, the strength of expressiveness conditioning mostly determines ethical attitudes, indicating that if the students have a higher level of expressiveness, they probably less accept cheating behavior (Becker & Ulstad, 2007).The data obtained from PAQ showed that participants have a higher level of expressiveness rating (e.g. with a higher level of female ratings).Further, analysis obtained from ADS and the covariate analysis of the expressiveness (EXPRESS) of the students' gender is not significant (p>.01 but EXPRESS is a significant covariate in the analysis (p=.001), indicating that there is no gender effect after the inclusion the student expressiveness in the analysis.Finally, to investigate the students' gender difference in the ratings of cheating behavior if the impression management is considered, a covariate analysis was performed.The results indicated that the females have a lower level of impression management than the male have and gender and impression management do not have any significant effect on cheating behavior.A study by Witmer and Johanson (2015) showed that female students are less prevalent in disciplinary matters and female students deny academic dishonest behaviors more than male students.In addition, in another study carried out by Becker and Ulstad (2007) male students were determined to be engaged in significantly less impression management than the females.Although much research found that females present less dishonest behaviors there is still other studies that found the reverse under low-risk conditions (Leming, 1980).For example, in a study conducted by Singh, Kumra and Vinnicombe (2002) male participants were found to be more willing to engage in impression management than did females.On the other hand, there are other studies indicating that dishonesty is not related to conditions however it is linked to the students' individual attitudes (Nonis & Swift, 2001).Overall, there is still an ongoing discussion regarding the differences in terms of dishonest behaviors of females and males.
This study is limited to only EFL learners' behaviors who have not received a detailed instruction on academic ethic instead they attended courses like research techniques of which some contents are on the student orientation to the academic honesty.Under the light of limitations, it is implicated that detailed instruction on academic ethics should be implemented in BA programs from the first year of education.Also, this study has not investigated the main motives that lead students to cheat.Understanding the factors that underlie the tendency of the students to display dishonest academic behaviors might be helpful in taking preventive precautions and designing appropriate coping strategies by school administrators and university instructors.Especially in the context of Turkey, there is a growing need to develop and implement cohesive frameworks for dealing with cases of alleged plagiarism.That's why, future research might investigate main factors that cause students to cheat from both instructors' and students' perspectives and might search for practical approaches for avoiding plagiarism.This will yield a deeper understanding of the problem.

Table 1
shows that female participants are nearly four times more than males.The number of the participants in universities varies.