Back=20

1 article(s) will be saved.

To continue, in Internet Explorer, select = FILE=20 then SAVE AS from your browser's = toolbar=20 above. Be sure to save as a plain text file (.txt) or a = 'Web Page,=20 HTML only' file (.html). In Netscape, select=20 FILE then SAVE AS = from your=20 browser's toolbar above.

Record: 1

Title:: Test cheating in a rural college: = Studying the=20 importance of individual and situational factors.
Author(s):: Robinson, Erin, Department of = Sociology, Social=20 Work and Criminology, Morehead State University, Morehead, KY, = US
Amburgey,=20 Rita, Department of Sociology, Social Work and Criminology, Morehead = State=20 University, Morehead, KY, US
Swank, Eric, Department of Sociology, = Social=20 Work and Criminology, Morehead State University, Morehead, KY, = US
Faulker,=20 Cynthia, Department of Sociology, Social Work and Criminology, = Morehead State=20 University, Morehead, KY, US
Source:: College Student Journal, Vol 38(3), = Sep 2004.=20 pp. 380-395.
Publisher:: US: Project Innovation of = Mobile.
ISSN:: 0146-3934 (Print)
Language:: English
Keywords:: test cheating; rural college; = individual=20 factors; situational factors; students; independent = variables
Abstract:: There is a long line of studies on = academic=20 cheating. This literature has established a litany of different = individual and=20 contextual factors that seem to inspire student conformity of the = official=20 rules of test taking. This study draws on this literature by using = fourteen=20 independent variables to explain the cheating habits of students in a = Central=20 Appalachian university (the model includes demographic factors as well = as=20 interpretations of variables related to professors and peers). While = this=20 study uses many of the familiar predictors of cheating for urban = settings, it=20 is unique in that it studies the impact of these factors in a rural = campus. In=20 the end, this quantitative analysis of 118 students reveals that the = dynamics=20 behind cheating might be universal. While rural communities might = offer=20 different dynamics for some issues, the extent of cheating at this = campus=20 mirrored the rates of studies from many urban schools (almost four of = five=20 students cheated in some way). Likewise, the inferential statistics = also=20 followed typical patterns. Almost all of the independent variables = presented=20 significant results in the bivariate analysis and the variables of = Greek=20 membership, enjoyment of college, peer cheating and the fear of = punishment=20 remained significant in a multivariate regression. (PsycINFO Database = Record=20 (c) 2007 APA, all rights reserved)(from the journal = abstract)
Subjects:: *Cheating; *Colleges; *Independent = Variables;=20 *Risk Factors; *Rural Environments; College Students
Classification:: Classroom Dynamics & Student = Adjustment=20 & Attitudes (3560)
Population:: Human (10)
Male (30)
Female=20 (40)
Location:: US
Age Group:: Adulthood (18 yrs & older) = (300)
Methodology:: Empirical Study; Quantitative = Study
Publication = Type:: Journal, Peer Reviewed Journal;=20 Print
Format(s) Available: Print
Document Type:: Journal Article
Release Date:: 20041025
Accession = Number:: 2004-19149-008
Number of Citations in=20 Source:: 43
Database:: PsycINFO
Full Text Database:=20: Academic Search Premier
Notes:: This title is available at = Camden-Carroll=20 Library.

TEST CHEATING IN A RURAL COLLEGE: STUDYING = THE=20 IMPORTANCE OF INDIVIDUAL AND SITUATIONAL FACTORS

There is a long line of studies on academic = cheating.=20 This literature has established a litany of different individual and = contextual=20 factors that seem to inspire student conformity of the official rules of = test=20 taking. This study draws on this literature by using fourteen = independent=20 variables to explain the cheating habits of students in a Central = Appalachian=20 university (the model includes demographic factors as well as = interpretations of=20 variables related to professors and peers). While this study uses many = of the=20 familiar predictors of cheating for urban settings, it is unique in that = it=20 studies the impact of these factors in a rural campus. In the end, this=20 quantitative analysis of 118 students reveals that the dynamics behind = cheating=20 might be universal. While rural communities might offer different = dynamics for=20 some issues, the extent of cheating at this campus mirrored the rates of = studies=20 from many urban schools (almost four of five students cheated in some = way).=20 Likewise, the inferential statistics also followed typical patterns. = Almost all=20 of the independent variables presented significant results in the = bivariate=20 analysis and the variables of Greek membership, enjoyment of college, = peer=20 cheating and the fear of punishment remained significant in a = multivariate=20 regression.

Academic dishonesty has probably existed since = the=20 inception of colleges. In turn, essays on student cheating have been = around for=20 at least the past 60 years (e.g., Groves, 1936). More recently, many = popular and=20 scholarly treatises have warned cheating has reached epidemic levels or = that=20 student cheating is inherent to the college experience (Michaels & = Miethe,=20 1989; Whitely, 1998).

This paper concentrates on the factors that = are=20 associated with cheating on tests. In using a term like cheating, we = mean the=20 "intentional use or attempted use of unauthorized materials, = information, or=20 study aids" in any work submitted for academic credit (Pavela, 1997). = With such=20 a definition, there is striking evidence that a large number of college = students=20 cheat. Recent studies have found that students, on anonymous surveys, = are quick=20 to admit to many sorts of academically dishonest deeds (varying from = giving=20 false excuses for missing deadlines to stealing copies of exams). Some=20 researchers suggest that about 50% of college students confess to = cheating at=20 some point during their academic careers (Whitley, 1998). Other studies = suggest=20 that cheating is even more rampant. For example, McCabe and Trevino's = (1996)=20 study of 6,000 students at 31 selective colleges and universities found = that 70%=20 of students admitted to cheating on exams, 84% to cheating on written=20 assignments, and almost half to inappropriately collaborating with = others on=20 assignments.

While studies on collegiate cheating have = flourished,=20 the type of schools sampled has been quite narrow. Almost all of the=20 quantitative studies on cheating have occurred at large and medium sized = universities that are in or close to major metropolitan centers. = Conversely, the=20 literature has provided few systematic explorations into the cheating = dynamics=20 of students from rural college campuses (see Graham et. al., 1994; Tang = &=20 Zuo, 1997; Cochoran et.al., 1999). In filling this research gap, this = work=20 explores the academic cheating practices of students from a rural = college in=20 Central Appalachia. In the descriptive analysis, the study elaborates = the extent=20 of different sorts of test cheating at this college. Later, this piece=20 identifies the factors that facilitate certain cheating practices. In = doing so,=20 this paper will determine if the variables that explain cheating in = urban=20 samples retain their predictive strength for students who matriculate at = a rural=20 college setting.

Literature Review

Researchers have identified a number of = situational,=20 dispostional and demographic factors that drive students to cheat. While = some of=20 this literature has yielded mixed and contradictory results, this study = sticks=20 with some of the variables that have withstood many tests of replication = (i.e.,=20 statistical significance in several studies).

In connecting cheating tendencies to = individualistic=20 traits, much of the research has tried to create profiles of cheaters = (Whitley,=20 1998). In studies of demographic backgrounds, most studies have found = that=20 cheating practices are equally distributed among peoples from different = races or=20 social-classes (Celabrese amp; Cochoran, 1990; Tang & Zuo, 1997; = Roig &=20 Ballew, 1994; Cochoran et.al., 1999). On the other hand, some studies = have found=20 a relationship between gender and academic dishonesty (Leming, 1980; = Newstead=20 et. al., 1996; McCabe & Trevino, 1997; Crown & Spiller, 1998; = Whitley,=20 1998). That is, studies repeatedly argue that male students were = slightly more=20 likely to cheat than their female counterparts (Ward, 1986; Tang & = Zuo 1997;=20 Michaels & Miethe, 1989; Celabrese & Cochoran, 1990; Davis & = Ludvigson, 1995; Genereux & McLeod, 1995; Hendershott, Drinan & = Cross,=20 1999; Tibbetts, 1999; Storch & Storch, 2002). Other studies link = academic=20 cheating to matters of age. Younger students seem to cheat more often = (Graham=20 et. al., 1994; Diekhoff et. al., 1996; Tibbetts, 1999; McCabe & = Trevino,=20 1997; Cochoran et.al 1999) as are first and second year students who are = just=20 beginning their college careers (Michaels & Miethe, 1989; Newstead = et. al.,=20 1996; Crown & Spiller, 1998; Jordan 2001).

While numerous studies connect academic = cheating to=20 individual attributes, other research suggests that certain social = contexts=20 inspire cheating (LaBeff et. al., 1990). In challenging the notion that = cheating=20 emerges from the qualities that rest within the students themselves, = this=20 position contends that certain conditions inspire the enactment of = cheating. In=20 emphasizing the role of the social environment, this line of inquiry has = established the importance of student interpretations of professorial = and=20 student cues.

Numerous studies have found the certain = impressions of=20 professorial conduct can augment or dampen student cheating. In a very = obvious=20 conclusion, it is clear that students cheat less when professors show a = real=20 concern over cheating (Roig & Ballew, 1994; Davis & Ludvigson, = 1995;=20 Genereux & McLeod, 1995). However, this factor by itself may be a = necessary=20 but not sufficient deterrent to cheating. That is, professorial comments = about=20 the immorality of cheating may have little effect since students can = ignore such=20 messages. Subsequently researchers have turned to other factors that = might=20 modify or alter the effects of these "just say no to cheating" = edicts.

Pertinent teacher cues can be connected to = matters of=20 class content. Student cheating seems to drop when professors are = enthusiastic=20 about the curriculum and students internalize an emphasis on the = intrinsic worth=20 of learning (Celabrese & Cochoran, 1990; Roig & Neawman, 1994; = Genereux=20 & McLeod, 1995). Converlsy, students are more likely to cheat when = they=20 think the class content is trivial, boring or irrelevant (LaBeff et. = al., 1990;=20 Genereux & McLeod, 1995; Pulvers & Diekhoff, 1999). In a similar = fashion, students who prioritize the notion of understanding class = information=20 are less likely to cheat than the students who focus on getting good = grades or=20 simply passing a required class (Huss et. al., 1993; Diekhoff et. al., = 1996;=20 Newstead et. al., 1996; Jordan, 2001). Similarly, students who primarily = attend=20 college for extrinsic purposes such as "getting a better job" or to = "impressing=20 future employers" are more inclined to cheat (Michaels & Miethe, = 1989; Davis=20 & Ludvigson, 1995; Jordan, 2001).

As perceptions of content matters, so do = impressions of=20 faculty mannerisms. In general terms, students who like and respect the=20 interaction style of their teachers are less likely to cheat (LaBeff et. = al.,=20 1990; Diekhoff et. al., 1996; Hendershott, Drinan & Cross, 1999; = Pulvers=20 & Diekhoff, 1999; Steams 2001). Accordingly, cheating normally = decreases=20 when teachers are seen as fair graders and it skyrockets when students = think=20 professors deliberately make unreasonably tough tests (Celabrese & = Cochoran,=20 1990; LaBeffet. al., 1990; Graham et. al., 1994; Davis & Ludvigson, = 1995;=20 Genereux & McLeod, 1995; McCabe & Trevino, 1997). Similarly, = impressions=20 that teachers are kind, respectful and understanding will lessen = cheating while=20 the opposite is true when professors are seen as being grumpy, callous = or=20 unfriendly (LaBeff et.al, 1990; Genereux & McLeod, 1995; Waugh et. = al.,=20 1995; Diekhoff et. al., 1996; Ashworth & Bannister, 1997).

Another component of cheating can be notions = of how=20 people achieve good grades. While it is possible that students who do = well in=20 college are simply good cheaters, the research suggests otherwise. While = the=20 pressure to get good grades might temp all sorts of students, it is the = students=20 with the greatest academic difficulties who are more prone to act on = these=20 urges. In using grade point average as an indicator of academic = achievement,=20 study after study have found a significant negative relationship between = cheating and GPA (Antion & Michael, 1983; Tang & Zuo 1997; = Graham et.=20 al., 1994; Roig & Neaman, 1994; Davis & Ludvigson, 1995; = Diekhoff et.=20 al., 1996; McCabe & Trevino, 1997; Crown & Spiller, 1998; = Tibbetts,=20 1999). On top of "objective measures" of academic achievement, = subjective=20 measures of academic confidence also report significant results. Both = experiment=20 and field surveys have found that students are more likely to cheat when = they=20 doubt their intelligence, lack academic confidence or expect failure = (Leming,=20 1980; Ward, 1986; LaBeffet. al., 1990; Schab, 1991; Tang & Zuo = 1997). Others=20 studies add that cheaters often worried that that they cannot reach = their=20 professors standards for performance (Barnett & Dalton, 1981; Davis = &=20 Ludvigson, 1995) and this fear leads many to procrastinate in their = studies=20 (Michaels & Miethe, 1989; Davis & Ludvigoson, 1995). In turn, = this=20 procrastination can lead to more cheating since students who study less = cheat=20 more often (Michaels & Miethe, 1989; Whitely, 1998).

In trying to escape bad grades, many students = turn to=20 competitive and individualistic solutions. In turn, cheating is more = common=20 among those students who believe that one reaches good grades through = the means=20 of individualistic striving and aggressive battles of one ups-manship = (Barnett=20 & Dalton, 1981; Perry et. al., 1990; Pulvers & Diekhoff, = 1999).

The implementation of cheating-specific = policies can=20 also alter cheating practices. Deterrence theories suggest that cheating = occurs=20 within the boundaries of expected costs and benefits (McCabe & = Trevino,=20 1997; Tibbetts, 1999). With a rational choice premise, this position = argues that=20 a perception of severe penalties dampens the likelihood of cheating. Or = put=20 otherwise, cheating behaviors are "inhibited or deterred in direct = proportion to=20 perceived probability and severity of punishment" (Michaels & Miethe = 1989,=20 p. 871). Early studies that used this theory suggested that the = opportunities to=20 cheat and the fear of external sanctions were the most significant = factor in=20 reducing cheating (Tittle & Rowe, 1974; Leming, 1980; Michaels & = Miethe,=20 1989) while later works argue that the fear of punishment is but one of = many=20 factors that decreases the amount of cheating (Graham et. al., 1994; = Genereux=20 & McLeod, 1995; Diekhoff et. al., 1996; McCabe & Trevino, 1997;=20 Hendershott, Drinan & Cross, 1999). Likewise, the effective = communication of=20 cheating policies combined with increased student awareness of common = penalties=20 tends to reduce dishonest behavior (Aaron, 1992; Genereux & McLeod, = 1995;=20 Crown & Spiller, 1998, Jordan, 2001).

Student habits also develop in the context of = other=20 student cues. In taking the actions of other students as examples, = personal=20 closeness to rule-followers generally inspires conformity to the = official rules=20 of test taking. With this being the case, several studies have found = that=20 students felt more comfortable with cheating when they believed that = their=20 friends tolerated such actions (Roig & Ballew, 1994; Diekhoff et. = al.,=20 1996). Similarly, the perceived cheating behavior of the general student = populace also seems to matter. Several studies have found that students = are more=20 likely to cheat if they think that cheating is widespread among all of = the=20 students in their college (Graham et. al., 1994; Genereux & McLeod, = 1995;=20 McCabe & Trevino, 1997; Hendershott, Drinan & Cross, 1999; = Jordan, 2001)=20 or a majority of the students in their classrooms (Genereux & = McLeod, 1995;=20 Diekhoff et. al., 1996). Likewise, living with cheating roommates = increases=20 cheating behaviors (Jordan, 2001) as does the simple act of living in = college=20 residence halls (Graham et. al., 1994). Finally, studies find the = physical act=20 of watching others cheat presents even larger results. That is, some of = the=20 strongest correlations transpire when students literally observe their = peers=20 successfully cheating (Diekhoff et. al., 1996; Crown & Spiller, = 1998).

While the actions of the general student = populace should=20 sway cheating, membership in certain primary groups can present larger = impacts.=20 When comparing the influence of close school friend and casual = acquaintances,=20 several studies have concluded that the encouragement from one's friend = has a=20 greater impact (Michaels & Miethe, 1989; Tibbetts, 1999; Jordan, = 2001).=20 Additionally, other works concluded that direct pressure from one's = friend to=20 cheat can have an enormous impact on cheating (LaBeff et. al., 1990; = Genereux=20 & McLeod, 1995; Diekhoff et. al., 1996; McCabe & Bowers, 1996). = Other=20 studies suggest that memberships in specific voluntary organizations can = exasperate certain cheating tendencies. Students who join fraternities = and=20 sororities show a greater propensity to cheat (Diekhoff et. al., 1996; = McCabe=20 & Bowers, 1996; McCabe & Trevino, 1997; Crown & Spiller, = 1998;=20 Whitely, 1998; Storch & Storch, 2002). It is thought that Greek=20 organizations provide increased opportunities for collusion and a pool = of=20 friends who are more likely to condone cheating (McCabe & Bowers, = 1996).

This literature review suggests that the = propensity to=20 cheat can spring from many sources. Since cheating can emerge from a = confluence=20 of different causes, the following analysis will focus on demographic = factors=20 (gender, age, class standing in college, greek membership, campus = residency,=20 grade point average), teacher-student interactions (professor demands, = ease of=20 cheating), student-student interactions (extent of peer cheating) and=20 educational attitudes (pleasures in learning, acceptance of competition, = purpose=20 of attending college, faith in their own academic capabilities, desire = to study=20 for tests). The rest of this paper will explore how these independent = variables=20 fared in explaining the test cheating practices of students at a rural=20 college.

Methods

Sampling

Information for this study was collected at a = public=20 university in Eastern Kentucky. Being a regional university in Central=20 Appalachia, the sampling unit has some unique characteristics. With most = students commuting from close locales, the undergraduate population = mirrors the=20 qualities of the surrounding communities. This means the student body is = predominantly white (95.2% of the 8,000 students are Euro-American), = many of the=20 students are first generation college students (only 7% of surrounding = counties=20 have adults who have bachelor's degrees) and there is noticeable = contingency of=20 older returning students (24% of the students are over 25 years old). = Moreover,=20 being in a poorer section of the United States, a large percentage of = the=20 students grew up in economically distressed communities. The = university's most=20 serviced counties have per capita in comes around $11,000, and poverty = rates=20 above 35% (Rural Development Working Group, 1995).

Respondents were selected through a purposive = sample.=20 Rather than conveniently distributing surveys to a single classroom, = surveys=20 were initially distributed to individuals who were enrolled in a diverse = set of=20 academic disciplines (social work, sociology, history, education,=20 communications, math, biology). The reason for selecting different = disciplines=20 is that some works suggest that cheating practices can vary by major and = academic department (Roig & Ballew, 1994; Genereux & McLeod, = 1995;=20 Newstead et. al., 1996). We also passed out questionnaires at the = student=20 athletic and dining centers on campus. These settings were chosen since = they=20 contain a higher density of greek affiliates and offer access to = students who=20 fail to attend their official classes.

The study acquired 118 competed surveys = through this=20 approach. Due to our screening procedures, all participants were = currently=20 enrolled as full time students. Fitting the campus profile, most of our = students=20 were white (89 Anglo-Europeans, 27 African-American, 2 Asian-Americans) = and the=20 majority female (65 women and 53 men). The sample also had almost equal=20 representation of students from different class standings (35 freshman, = 19=20 sophomores, 29 juniors and 35 seniors) and people who resided in campus = dorms=20 (53 of the participants lived on campus and 65 presently lived off = campus).

Measurement

In creating measures, the researchers crafted = a 45-item=20 survey. The first section of the questionnaire dealt with the dependent = variable=20 of test cheating. To assess this variable, students read the prompt: "In = the=20 last academic year, how often have you engaged in the following = actions?" A list=20 of seven ways to cheat rested beneath this question (similar to Michaels = &=20 Miethe, 1989; McCabe & Trevino, 1997). Some of the cheating styles = involved=20 working by oneself (i.e., sneaking illicit notes into a test) while = other items=20 dealt with group cheating techniques (getting answers from a student who = had=20 completed the test). In responding to such topics, students were asked = to=20 summarize their actions on a 4-point frequency scale (see Tables 1 for = exact=20 item and response wording). The response "never" was coded as 1, "once" = received=20 a 2, "two to ten times" met a 3, and "more than 10 times" found a 4. A = composite=20 measures was constructed through the addition of these items and the = scale=20 maintained a high level of parallel reliability (Cronbach Alpha=3D = .84).

Most of the independent variables were = addressed thought=20 the Likert Scale (a statement followed by five responses that went from = Strongly=20 Agree to Strongly Disagree). Some of the variables were ascertained = through a=20 single item. The variable unfair professor expectations was assessed by: = "Professors demand too much from their students at MSU" while the = perceived ease=20 of cheating was handled by: "I can cheat and get away with it" (SA=3D5). = The=20 variable student competitiveness was explored by: "I want to be brighter = than=20 other students" and vocational school motives read: "I mostly come to = college to=20 find a good job." The factor academic confidence was matched with the = comment:=20 "I am a pretty bright person" (SA=3D5).

Other more complicated variables had = multi-item indexes.=20 Like Michaels & Miethe (1989) and Jordan (2001), our peer cheating = index=20 dealt with habits of the general student body and the practices of = student=20 friends (Cronbach Alpha of .71). The specific items read: "Many of my = friends=20 cheat at college" and "Most students cheat at MSU." The student's = enjoyment of=20 college was assigned three phrases: "I find college interesting," "I = enjoy=20 reading," and "having intellectual conversations are fun" (Cronbach = Alpha of=20 .68).

The last section of the survey dealt with = demographics.=20 Dummy variables were created for the responses to "What is your gender? = .... do=20 you belong to a fraternity or sorority?" and "do you live on campus?" = (Females=3D=20 1, Males=3DO; Greek=3Dl, Other=3DO; On Campus=3Dl, Other=3DO). Class = standing was=20 identified by the categories of freshmen, sophomore, junior, and senior = (i.e.=20 freshman=3D1, seniors =3D4). The item for hours studying was: "For the = last academic=20 year, how many hours do you approximately spend studying a week?" = (none=3Dl, one=20 to five hours=3D2, six to ten hours=3D3, more than eleven=3D4). Finally, = there were=20 some open-ended questions. For academic achievement and age students = responded=20 to the questions: "What is your overall GPA at Morehead State?" and = "What is=20 your age?"

Findings=20

Descriptive Results

The extent of cheating is presented in Table = 1. Overall,=20 it seems that a majority of the students cheat on an occasional or = regular=20 basis. Most students have tried all but one of these actions with only = the=20 riskier act of sneaking cheat-sheets into the classroom presents a = majority of=20 abstainers (66% said they never did such). Conversely, the percentage of = students who admit to other cheating techniques ranges between 57 to 89 = percent=20 of the students. This means, about three out of five students have = copied from=20 other student exams or made their test available to other students. = Moreover,=20 three out of four students have received questions from someone who has=20 completed the test and an overwhelming nine out of ten students have = improperly=20 collaborated on a take home exam.

While most students have cheated some time = during their=20 college career, the propensity is not constant for the whole sample. A = segment=20 of the sample has had a single descent into all of these cheating = approaches.=20 That is, anywhere from 11 to 23% of the students tried these maneuvers = only=20 once. A larger faction of students seems to draw upon cheating routines = more=20 frequently. In practicing intermittent cheating routines, between 33 and = 70=20 percent of the respondents admit to using each technique between two and = ten=20 times. In turn, this high level of responses for these periodic bouts of = cheating means that the twice to ten times instances was the modal score = for=20 three items (exam copying without the student's knowledge, getting = answers from=20 other students, sharing notes in a take-home exam) and a near modal = score for=20 three other (exam copying with student's knowledge, helping another = cheat on a=20 test, offer answer cheats to others). Finally, the extent of chronic = cheating is=20 much rarer. Between one to four percent of the students admit to the 10 = plus=20 range and only one item meets the double- digit realm.

Explanatory Results

To identify the factors that inspire cheating, = we=20 utilized two types of statistical tests (see table 2). First, simple=20 correlations were used as a way to establish significant bivariate=20 relationships. Next, to control for issues of confounding variables and=20 spuriousness, a single OLS regression was ran for all of the independent = variables in the entire model (see column 2 of Table 2 for the Pearson=20 Coefficients and column 3 for the Standardized Beta Coefficients).

When exploring bivariate associations, eight = of the=20 fourteen variables presented statistical significance. Of the four = variables=20 that surpassed at the .001 alpha, ease of cheating showed a moderately = strong=20 relationships of .491 while peer cheating, enjoyment of college, greek=20 membership displayed moderate associations of .378 to .301. Four other=20 significant variables rejected the null at the .05 alpha as they offered = weak=20 relationships of. 187 to. 167 (Professor demands, vocational motivation, = gpa and=20 gender). Finally, the variables of college classification and hours = studying=20 almost reached a significant level while issues of age, competitiveness = and=20 academic confidence presented minimal associations.

While many of these variables were significant = in a=20 bivariate analysis, issues of colinearity and mitigating factors led us = to a=20 multivariate analysis. Even when controlling for the influences of the = other=20 factors, four of the variables retained their statistical significance = in an OLS=20 regression (see column 3 for beta coefficients). The ease of cheating = presented=20 the highest beta coefficient with a moderate direct relationship of .332 = while=20 greek membership presented a respectable beta weight of .213. Two other=20 variables retained their significance but presented less of an impact in = the=20 multivariate context. In presenting standardized scores around. 15, both = enjoyment of college and peer cheating were only able to reject the null = at the=20 .05 alpha. Other factors that were significant in bivariate calculations = dropped=20 completely out of the significance realm. Vocational motives, = perceptions of=20 teacher fairness, a student's GPA and gender did not present significant = direct=20 impact on the amount of test cheating in this sample (with a slightly = larger=20 sample, the vocational beta could have rejected the null). Thus, = cheating seems=20 to arise from a convergence of four forces. Test cheaters were inspired = by the=20 opportunity of a safe cheating experience, the absence of pleasure in = learning,=20 the membership in a Greek organization and the belief that other = students=20 support their cheating ways.

As several of the variables maintained = significance in=20 the regression, the entire theoretical model netted substantial results. = With a=20 R-squared of .395, this set of variables could explain about 40% of the = variance=20 in the dependent variable. While this means that other antecedents must = induce=20 some of the cheating process, this score is somewhat impressive in the = circles=20 of the social and behavioral sciences (F-Score of 5.225 is significant = at the=20 .001 level).

Discussion=20

This study intends to introduce a greater = understanding=20 of cheating on rural college campuses. While anticipating some unique = qualities=20 of this population, we discovered some cheating frequencies that closely = parrelled the cheating proclivities of city-based college students. = Consistent=20 with the studies on urban undergraduates, the descriptive figures = suggest that=20 occasional cheating is the norm. While the use of secret cheat sheets = were=20 rejected by most students in this sample, anywhere between 37 to 78% of = the=20 students used six of the seven cheating techniques more than once. On = the higher=20 end, three out of four students regularly shared notes on a take home = exam and=20 three out of five of students have occasionally acquired answers from = someone=20 who has already completed the test. The other actions of helping others = on a=20 test and in-class exam copying were slightly rarer events, but almost = half of=20 the respondents have used these techniques at least twice.

With cheating being the norm, our analysis = also sought=20 to identify the factors that inspire or dampen academic dishonesty. The = simple=20 correlations offered some early insights. Most of the variables that = highlighted=20 individualistic and psychological characteristics never reached = statistical=20 significance. That is, neither age, class standing, hours of studying = and campus=20 living did not significantly bared upon cheating practices, nor did the=20 attitudinal predispositions of favoring competitiveness or displaying = confidence=20 in one's intellectual skills. In turn, gender was the only social status = that=20 showed any consistent patterns (being a women seems to curb cheating = practices).=20 With most of the personal qualities displaying little impact in = corelational=20 analysis, the contextual and situational factors carried much greater = weight.=20 Professor-student interactions seemed to matter in bivariate conditions = since=20 students were more likely to cheat if they believed their professors set = too=20 high of standards or if the instructor seemed unable to catch their = cheating.=20 Likewise, cheating increased when college was deemed tedious and when = students=20 prioritized the careerist side of college. In addition, interactions = with=20 students were important. Students were more likely to cheat when they = socialized=20 with student cheaters and when they joined Greek organizations. Thus, in = synthesizing these bivariate findings, the rural test cheater is most = likely to=20 be a male fraternity member who thinks cheating is rampant as he looks = for safe=20 chances to practice his cheating. In addition, he complains about = working too=20 hard as he expects his college education to provide him with a hefty = salary=20 after graduation.

In moving to multivariate calculations, four = of the=20 independent variables met the minimal threshold for statistical = significance. In=20 identifying these factors, this work suggests that cheating emerged from = a=20 mixture of meanings that were attached to both teacher and student = expectations.=20 When addressing student-teacher interactions, cheating seemed to have an = opportunistic quality. That is, cheating was not a matter of right or = wrong, but=20 rather a matter of risk assessment. This means that cheating was = restrained when=20 there was a fear that cheating would be detected and punished. = Conversely, a=20 sense that cheaters can "win" and avoid negative consequences only seems = to=20 encourage wider acts of cheating. This study also showed that cheating = also=20 hinges on the desires behind student learning. Disillusioned or cynical = students=20 who find little joy in reading or intellectual exchanges were quicker to = rely on=20 cheating. Conversely, cheating was rare among enthusiastic students who = sought=20 the intrinsic side of learning. Thus, in the minds of students, cheating = was=20 more acceptable when professors do not properly perform their roles as=20 entertainers and enforcers. Or in other words, cheating is considered = less=20 reprehensible or even legitimate when a class is deemed "boring" or if = the=20 professor seems "ill-suited" to stop cheating.

While cheating practices seemed contingent = upon=20 perceptions of teacher conduct, the framing of peer etiquette was = equally=20 important. The perceived endorsement of cheating by fellow students = seems to=20 induce more cheating. In essence, test cheaters were convinced that most = of=20 their friends and student acquaintances tolerated or condoned academic = misdeeds.=20 Moreover, cheaters thought that student cheating is pervasive and that = their=20 actions were simply conforming to a larger pattern. With such an = impression,=20 cheating appeared proper or reasonable since it was considered typical = of the=20 college experience. Finally, belonging to a fraternity presented another = direct=20 effect. Thus, regardless of how one views teacher and student = interactions, the=20 membership in a greek organization dramatically increased the likelihood = of=20 cheating.

While four variables presented crucial = contributions in=20 the multivariate analysis, the rest did not. Six of the variables were = not=20 significant in any of the statistical calculations (academic confidence, = age,=20 class standing, competitiveness, on-campus living, hours studying). Four = of the=20 variables were significant only in the initial bivariate computations. = Similar=20 to other studies, the individual factors of gender and GPA did not = retain their=20 significance in an OLS regression (Tang & Zuo 1997; Cochoran et.al., = 1999;=20 Jordan, 2001). Likewise, both the claims of excessive academic rigor and = an=20 adherence to vocational ambitions also lost their salience in the = multivariate=20 context. Subsequently, while cheating was more common among vocationally = orientated males who combined low grades with gripes about the = difficulty of=20 class assignments, these variables do not directly drive cheating. = Instead,=20 students who hold these characteristics are more likely to retain the = factors=20 that really inspired cheating (trivializing the worth of intrinsic = learning,=20 having confidence in their cheating abilities, joining fraternities and=20 perceiving widespread cheating among other students). Thus, this work = supports=20 the assertion that women tend to be more intrinsically motivated in = college=20 (Newstead et. al., 1996) and that students who come to college for = employment=20 purposes are drawn to fraternities and other students who cheat = (Diekhoff et.=20 al., 1996). This work also illustrates the falsity behind the conclusion = that=20 receiving low grades automatically induces cheating. Instead, what = really seems=20 to matter is the way in which students make attributions for their bad=20 grades.

Limitations

Before addressing the ramifications of this = research, we=20 hope that future research builds upon this initial analysis. While our=20 theoretical model presented some robust findings, it is clear that the = study has=20 not captured every crucial factor. Future researchers might integrate = new=20 variables into the mix since such an inclusion could delineate more = variable=20 interactions or add to the potency of the total model (for an exhaustive = review=20 of the literature see Whitely 1998).

Along methodological lines, future works might = improve=20 on our research design. The use of a cross-sectional research design = means that=20 this study cannot elaborate the temporal order of events. In turn, this = approach=20 can make claims over causation problematic. For example, it is = impossible to=20 know if teacher impressions preceded the act of cheating. Likewise, the = data is=20 unable to determine if students became cheaters after watching other = student=20 behaviors of if academically dishonest sought out friends who reinforce = their=20 pro-cheating tendencies (readers will have to flip the logic of this = study's=20 independent and dependent variables if either of these scenarios are = correct).=20 To confront these dilemmas, future researchers might develop some = longitudinal=20 studies that contain "objective" measures of friendship networks and = professors=20 performances. In doing so, researchers will be able to distinguish if = cheating=20 really is a response to lackluster professors or the discovery that = others cheat=20 in their college environment. Conversely, such an approach could = determine if=20 most of these student claims are simply post hoc justifications that = follow=20 earlier decisions to cheat (Michaels & Miethe, 1989; Graham et. al., = 1994;=20 Cochoran et. al., 1999; Tibbetts, 1999).

The use of surveys also carries inherent = shortcomings.=20 There are always concerns over social desirability and measures are = never=20 perfect. For example, our measures probably underestimated the extent of = cheating since students may feel uncomfortable admitting to actions that = are=20 suppose to be improper (to lessen this problem, the answers were = anonymous and=20 the surveys were distributed by the student members of this research = team).=20 Moreover, some of the single item measures might be stronger with a = full-scale=20 index or different coding procedures. For example, the variable academic = confidence might have carried stronger explanatory power if we included = some=20 standardized psychometric scales on self-esteem or if our categories = made=20 clearer distinctions between students who have high or moderate levels = of=20 expected success (Houston, 1978).

In matters of sampling, there might be two = concerns.=20 With a larger sample size, more of the independent variable could have = reached=20 statistical significance. On issues of representativeness, the behaviors = of=20 Appalachian students may not totally reflect the attitudes of all rural = college=20 students (public or private colleges from other rural regions might = enroll a=20 different sort of students). Thus, ensuing research might enlist = information=20 from campuses that are distributed throughout the entire country. On = other hand,=20 fears over generalizability and representativeness may be overstated = since the=20 descriptive and explanatory findings of this work mirrors the results of = previous multicampus studies (i.e., Genereux & McLeod, 1995; McCabe = &=20 Trevino, 1997).

Implications

Regardless of these methodological = limitations, this=20 work offers some insights for educational planners and practitioners. In = building policies from this research administrators and faculty can take = several=20 sorts of approaches. First, institutions can reconceptualize their = assumptions=20 of proper test parameters. Cheating would disappear if colleges = abolished the=20 whole test taking process. However, this approach might encounter = resistance=20 since current political climate seems enamored with the wider use of=20 standardized tests and educational outcome measures. Less transformative = approaches could change the definitions of how tests are conducted. In = modifying=20 the rules of the game, teachers may allow the use of notes during exams = or let=20 students share information. Again, institutions and professors may = reject such=20 alterations since they may challenge their assumptions of what = constitutes a=20 test (i.e., these test modes emphasize less memorization and strives for = greater=20 student cooperation).

If universities preserve the traditional = definitions of=20 test taking, smaller incremental changes can be enacted. Colleges can = try to=20 attract students who are less likely to cheat. Recruiting efforts can = focus on=20 finding students who enjoy the learning process and admission committees = might=20 turn away students who attend college for primarily monetary reasons. = Similarly,=20 applicants who cheated during high school might be rejected since many = studies=20 find a continuity between high school and college cheating (Sims, 1993; = Whitely,=20 1998). However, since the authors are drawn to the democratic principles = of=20 equal access to college, we recommend that colleges focus more of their=20 attention on altering the college environment. Colleges can create = smaller=20 classroom sizes since such actions can lessen cheating (Houston, 1986; = Whitely,=20 1998) or they can reform or eliminate the Greek system (see the = literature on=20 student substance abuse, sexual violence and racial antagonism for = further=20 reasons to disband the Greek system).

Colleges can rely on other venues to influence = student=20 culture. At a bare minimum, university handbooks and catalogues should=20 explicitly elaborate the school's position on academic integrity. These=20 statements in official documents should be supplemented with discussions = of=20 academic honesty in actual classrooms. Moreover, colleges might = establish a=20 student honor code since some studies suggest that the creation of these = agreements can slightly lessen cheating (Mays & Lloyd, 1993; McCabe = &=20 Trevino, 1993; Whitely, 1998). Colleges can also offer or expand on the = types of=20 programs and interventions that address the issues of studying skills,=20 procrastination and test anxiety (see Hattie et. al., 1996).

In the classroom, teachers should continue = their efforts=20 at being interesting and fair. For example, Stearns (2001) suggests that = cheating will lessen if teachers are more vocally expressive, smile more = often,=20 use more eye contact and present a welcoming body posture. Along the = lines of=20 academic rigor, instructors could simplify class assignments to the = point that=20 every student will feel capable of success. However, this decision might = run=20 counter to the mission of academic institutions. Instead, it seems = better to=20 explain one's rationale behind the difficulty of the tasks (i.e., good = writing=20 requires much editing or conducting original research will improve their = performance at work or graduate school). While students might not agree = with=20 these assessments, such elaborations will probably leave the impression = that the=20 teacher's rigor is done in the student's best interest.

Other works have pointed to specific actions = that can be=20 done during a test. With cheating being linked to cost-benefit = calculations,=20 creating an impression of vigilance might discourage some cheating. To = do this,=20 some researchers have suggested the techniques of utilizing several = proctors,=20 placing students far apart, using essay exams, employing different = versions of=20 tests, checking picture IDs, and collecting notebooks before a test = (Hollinger=20 & Lanza-Kaduce, 1996; Diekhoffet. al, 1996; Sterns, 2001). However, = teachers=20 should realize that most of these measures have minimal effects on = student=20 cheating attitudes or behaviors (Hollinger & Lanza-Kaduce, 1996). = Moreover,=20 an overly zealous or vindictive implementation of such procedures can = result in=20 counterproductive outcomes. First off, such actions can leave a = confrontational=20 or distrustful undertone to class dynamics. In turn, learning can lose = all of=20 its intrinsic worth when professors are seen as domineering enforcers = who=20 rigidly adhere to petty rules. Thus we suggest that the major point of=20 intervention should not be at the point of actually taking a test, but = rather=20 the entire college community must alter its larger macro practices that = create a=20 general predisposition to cheating.

Table 1 Descriptive Statistics for Test Cheating =

Legend for Chart:

A - Item
B - Never
C - Once
D - 2 to 10
E - 10 or more
F - Mean
G - Standard Deviation

             A                 B      C       D
                               E      F       G

Copying from another
student during an exam
"without his/her
knowledge.                    37%     21%     40%
                               2%   2.059   0.918

Copying from another
student during an exam
with his/her knowledge.       40%     23%     35%
                               2%   1.975   0.910

Using un-permitted notes
during an exam.               66%     18%     15%
                               1%   1.508   0.782

Getting questions or
answers from someone
who has already taken the
test                          24%     15%     50%
                              11%   2.483   0.976

Helping someone
else cheat on a test.         42%     17%     37%
                               3%   2.017   0.969

Shared notes from a take
home exam                     11%     11%     70%
                               8%   2.746   0.753

Made your answer sheet
available to enable another
student to see the answer.    43%     20%     33%
                               4%   1.983   0.969

Table 2 = Correlations=20 and Regression for Test Cheating

Legend for =
Chart:

A - Variable Name
B - Pearson R Score
C - Beta Coefficient

                 A                         B            C

Gender (Female=3D 1)                     -.180(*)      -.059
Age                                     .008          .004
College Classification (Freshman=3D1)    -.135         -.017
Unfair Professor Demands                .187(*)       .065
Ease of Cheating                        .491(****)    .332(***)
Enjoyment of College                   -.378(***)    -.154(*)
Student Competitiveness                 .080          .060
Vocational Motivation                   .166(*)       .120
Academic Confidence                    -.110         -.027
GPA                                    -.167(*)      -.073
Hours Studying                         -.145         -.032
Peer Cheating                           1329(***)     .156(*)
Greek Membership                        .301(***)     .213(**)
On Campus Living                       -.080         -.031
R-Squared                                             .395
F-Score                                              5.225(***)

Note: Beta coefficients are the standardized scores of the
regressions. p (*) <.05 p (**) <.01 p (***) <.001

Referenc= es=20

A= aron,=20 R.M. (1992). Student academic dishonesty: Are collegiate institutions = addressing=20 the issue? NASPA Journal, 29, 107-113

A= ntion,=20 D.L. & Michael, W.B. (1983). Short-term predictive validity of = demographic,=20 affective, personal, and cognitive variables in relation to two = criterion=20 measures of cheating behaviors. Educational and Psychological = Measurement, 43,=20 467-482.

A= shworth,=20 P. & Bannister, P. (1997). Guilty in whose eyes? University = students'=20 perceptions of cheating and plagiarism in academic settings. Studies in = Higher=20 Education, 22(2), 187-204.

B= arnett,=20 D. & Dalton, J. (1981). Why college students cheat. Journal of = Student=20 Personnel 22, 545-551.

C= elabrese,=20 R. & Cochoran, J. (1990). The relationship of alienation to cheating = among a=20 sample of American adolescents. Journal of Research and Development in=20 Education, 23 (2), 65-72.

C= ochoran,=20 J., Chamlin, M., Wood, P. & Sellers, C. (1999). Shame, embarrassment = and=20 formal sanction threats: Extending the deterrence-rational choice model = to=20 academic dishonesty. Sociological Inquiry, 69(1), 91-105.

C= rown,=20 D., & Spiller, S. (1998). Learning from the literature on collegiate = cheating. Journal of Business Ethics, 17(6), 687-700.

D= avis,=20 S. & Ludvigson, H. (1995). Additional data on academic dishonesty = and a=20 proposal for remediation. Teaching Psychology, 22(2), 199-222.

D= iekhoff,=20 G., LaBeff, E., Clark, R.,. Williams, L., Francis, B., & Haines, V. = (1996).=20 College cheating. Research in Higher Education, 37(4), 487-501.

= Genereux,=20 R.L., & McLeod, B.A. (1995). Circumstances surrounding cheating: A=20 questionnaire of college students. Research in Higher Education, 36 (6), = 687-704.

= Graham,=20 M., Mondray, J., O'Brien, K., & Steffen, S. (1994). Cheating at = small=20 colleges. Journal of College Student Development, 35, 255-260.

= Groves,=20 E.R. (1936). Let's face the problem of cheating. The National = Parent-Teacher=20 Magazine, pp.6-7, 24.

= Hattie,=20 J., Biggs J., & Purdie N. (1996). Effects of learning skills = interventions=20 on student learning: A meta-analysis. Review of Educational Research, 66 = (2),=20 99-136.

= Hindershott,=20 A., Drinan P., & Cross M. (1999). Gender and Academic Integrity. = Journal of=20 College Student Development, 40(4), 345-57.

= Hollinger,=20 R. & Lanza-Kaduce L. (1996). Academic dishonesty and the perceived=20 effectiveness of coutermeasures. NASPA Journal, 33(4), 292-306.

= Houston,=20 J. (1986). Survey corroboration of experimental findings on classroom = cheating=20 behavior. College Student Journal, 20 (2), 168-73.

= Houston,=20 J. (1978). Curvilinear relationships between anticipated success, = cheating=20 behaviors, temptations to cheat, and perceived instrumentality of = cheating.=20 Journal of Educational Psychology 75(4), 572-575.

= Huss,=20 M., Curnyn, J., Roberts S., Davis, S., Yandell, L. & Giordano, P. = (1993).=20 Hard driven but not dishonest. Bulletin of the Psychonomic Society, = 31(5),=20 429-430.

= Jordan,=20 A. (2001). College Student Cheating: The role of motivation, perceived = norms,=20 atttitudes and knowledge of institutional policy. Ethics and Behavior, = 11(3),=20 233-247.

= LaBeff,=20 E.E., Clark, R.E., Haines, V.J., & Diekhoff, G.M. (1990). = Situational ethics=20 and college student cheating. Sociological Inquiry, 60, = 191-198.

= Leming,=20 J. (1980). Cheating Behavior, subject varaibles and component of the=20 internal-external scale under high and low risk conditions. Journal of = Education=20 Research, 74(2), 83-87.

= May,=20 K. & Loyd B. (1993). Academic dishonesty: The honor system and = student=20 attitudes. Journal of College Student Development, 34(2), = 125-29.

= McCabe,=20 D. & Bowers, W. (1996). The relationship between student cheating = and=20 fraternity or sorority membership. NASPA Journal, 33(4), = 280-291.

= McCabe,=20 D. & Trevino, L. (1997). Individual and contextual influences on = academic=20 dishonesty: A multi-campus investigation. Research in Higher Education, = 38(3),=20 379-396.

= McCabe,=20 D. & Trevino, L. (1993). Academic dishonesty: Honor codes and other=20 contextual influences. Journal of Higher Education, pp. = 522-538.

= Michaels,=20 J. & Miethe, T. (1989). Applying theories of deviance to academic = cheating.=20 Social Science Quarterly, 70(4), 870-885.

= Newstead,=20 S., Franklyn-Stokes, A. & Armstead, P. (1996). Individual = differences in=20 student cheating. Journal of Educational Psychology, 48(2), = 229-241.

= Pavela,=20 G. (1997). Applying the power of association on campus: A model code of = academic=20 integrity. Journal of Business Ethics, 16(1), 97-118.

= Perry,=20 A., Kane,K., Bernesser, K. & Spicker, P. (1990). Type a behavior,=20 competetive achievement-striving and cheating among college students.=20 Psychological Reports, 66, 459-465.

= Pulvers,=20 K. & Diekhoff, G. (1999). The relationship between academic = dishonesty and=20 college classroom environment. Research in Higher Education, 40(4),=20 487-498.

= Roig,=20 M. & Ballew, C. (1994). Attitude toward cheating of self and others = by=20 college students and professors. Psychological Record, 44 (1), = 3-13.

= Roig,=20 M. & Neaman, M. (1994). Alienation, learning or grade orientation, = and=20 achievement as correlates of attitudes toward cheating. Perceptual and=20 MotorSkills, 78, 1096-1098.

= Rural=20 Devolpment Working Group. (1995). Socio-Economic Atlas of Kentucky. = College of=20 Agriculture: University of Kentucky.

= Schab,=20 F. (1991). Schooling without learning: Thirty years of cheating in high = school.=20 Adolescence 26 (104), 839-48.

= Sims,=20 R. (1993). The relationship between academic dishonesty and ethical = business=20 practices, Journal of Education for Business, 68(2), 207-211.

= Stearns,=20 S (2001). The Student-instructor relationship's effect on academic = integrity.=20 Ethics and Behavior, 11 (3), 275-286.

= Storch,=20 E. & Storch, J. (2002). Fraternities, sororities and academic = dishonesty.=20 College Student Journal, 36(2), 247-253.

= Tang,=20 S. & Zuo, Z. (1997). Profile of college examination cheaters. = College=20 Student Journal, 31(3), 340-347.

= Tibbetts,=20 S. (1999). Differences between men and women regarding decisions to = commit test=20 cheating. Research in Higher Education, 40 3), 323-341.

= Tittle=20 C.R., & Rowe A.R., (1974) Fear and the student cheater, Change, 6,=20 47-48.

= Ward,=20 D. (1986). Self-Esteem and dishonest behavior revisited. Journal of = Social=20 Psychology, 126, 709-713.

= Waugh,=20 R., Godfrey, J., Evans, E. & Craig D. (1995). Measuring student = perceptions=20 about cheating in six countries. Australian Journal of Psychology, 47,=20 73-80.

= Whitley,=20 B. (1998). Factors associated with cheating among college students. = Research in=20 Higher Education, 39, 235-274.

~~~~~~~~

By Erin Robinson, Department of Sociology, = Social Work=20 and Criminology Morehead State University; Rita Amburgey, Department of=20 Sociology, Social Work and Criminology Morehead State University; Eric = Swank,=20 Department of Sociology, Social Work and Criminology Morehead State = University=20 and Cynthia Faulkner, Department of Sociology, Social Work and = Criminology=20 Morehead State University

Copyright of College Student Journal is the = property=20 of Project Innovation, Inc. and its content may not be copied or emailed = to=20 multiple sites or posted to a listserv without the copyright holder's = express=20 written permission. However, users may print, download, or email = articles for=20 individual use.

Back=20