Tamera B Murdock & Anne S Beauchamp. 21st Century Education: A Reference Handbook. Editor: Thomas L Good. Sage Publications, 2008.
Attention to academic dishonesty (aka cheating) by the popular press has burgeoned recently as seen in John Stockwell’s movie Cheaters, Prime Time Live’s (Velmans & Koppel, 2004) special on cheating and numerous newspaper articles depicting tales of teachers trying unsuccessfully to curb cheating in their classrooms or losing their own jobs for altering students’ responses on standardized tests. This growing focus on cheating appears to parallel increases in academic dishonesty at all levels of education (Cizek, 1999). Moreover, students, teachers, and other school personnel have voiced concerns about cheating including the unfair advantage it provides to those who are dishonest and the ways in which it distorts the accuracy of available information about what students have actually learned. After defining academic cheating, this introduction presents an overview of the prevalence of cheating across age groups and discusses how students cheat.
What is Cheating?
Establishing a definition of academic dishonesty is complicated by numerous factors. Whereas some definitions focus solely on behavior, others specify information about the intent and consequences of the behavior. For example, one definition limits cheating to dishonest behavior that brings about an unfair advantage (Garavalia, Olson, Russell, & Christensen, 2007). Does that mean, for example, that Mark did not cheat if he receives a poorer exam grade than he might have achieved through his own efforts because he copied all the answers to the exam off Andre who, unbeknownst to Mark, had an alternate exam? If Tamara plagiarized on her history paper but did not intend to, has she cheated?
Cheating has been classified into four categories: the unapproved transfer of information between individuals, the use of unapproved materials, exploiting weaknesses in others, and plagiarism (Cizek, 2003). Although three of these four categories reflect behaviors, behavioral definitions are also problematic because of the lack of consensus as to some of the things that do and do not constitute cheating. For example, almost all university students and faculty agree that behaviors like copying off others on examinations, using crib sheets, and buying papers off the Internet to submit as one’s own are definitely cheating, there is less of a consensus about other behaviors such as copying homework, handing in the same paper for multiple classes, or working on projects with peers outside of class when the teacher has not specifically indicated it was all right to do so (Whitley & Keith-Spiegel, 2002). Disparities between teacher and student views of the seriousness of various behaviors have been reported at the middle and high school level (Evans & Craig, 1990). Further, Kohn (2007) points out, even within the same school, some teachers may encourage collaborative homework completion, whereas others would call it dishonest.
Across the definitional debates, the clearest common components of what constitutes academic cheating include the use or provision of unauthorized means of information in a setting where there are assessment consequences for the performance (Garavalia et al., 2007). This is the definition used in this chapter. It provides some clear, common components of academic dishonesty, but it also acknowledges that the specific behaviors that are considered to be cheating will vary across settings, and must be defined by the course instructor and/or the administration of the school. Behavior by teachers or other educational personnel who violate the procedures for the administration of standardized tests such as providing students with hints on problems or changing their answers before submitting them for scoring are also subsumed under this definition.
Prevalence of Cheating from Elementary School Through College
Prevalence estimates for cheating vary widely depending on the source of the information largely because of inconsistency in how cheating is defined. Here, we rely on published studies that have defined cheating behaviorally, with behaviors that are commonly agreed upon to constitute academically dishonest behavior: cheating on tests and copying papers or work from others.
Most research on academic cheating has been conducted with students in high school and college, thus providing more accurate prevalence estimates for these age groups. What little is known about cheating among younger children suggests that students start copying from each other as early as elementary school, and some students this age have already learned to use portable electronic devices such as cell phones and personal digital assistants (PDAs) to pass information to their peers during examinations. The most recent large-scale evidence of cheating in elementary school comes from a survey of over 1,000 sixth-grade students attending 45 elementary schools in California (Brandes, 1986). Over 85% of the students had seen cheating occur at least once and 30% stated seeing cheating happen many times.
From middle school forward, rates of academic dishonesty continue to increase. According to a nationally representative survey of 36,000 adolescents conducted by the Josephson Institute of Ethics (2006), approximately 38% of middle school students and 60% of high school students admitted to cheating on a test during the prior school year, with 14% and 35%, respectively, admitting to cheating two or more times. During the same time frame, copying Internet documents for classroom assignments was reported by approximately 24% of middle schools students and 33% of those in high school. Increases in both kinds of cheating were fairly linear from Grades 6-12. Across most years of the survey, there were no differences in rates of test cheating or plagiarism between students attending public schools and those attending private religiously orientated schools; however, there was less reported dishonesty among students from private nonreligious schools as compared to all other school types. Moreover, although rates of reported test cheating generally increased throughout the 1990s, the 2006 data actually reflect a decrease in reported cheating from the 1998 survey when 54% of middle school students and 70% of high school students said they had cheated on a test at least one time that year.
Although widespread cheating also occurs in college, most of the available data suggest there is less academic dishonesty in higher education than in high school, but that rates have increased in the past several decades. A comparison of two research reports conducted in 1963 and 1993 shows that whereas 26% of male college students said they copied on an exam and 16% used crib notes, and 30% plagiarized, in 1963, these numbers were 52%, 27%, and 26%, respectively, in 1993 suggesting increases in all categories except for plagiarism (McCabe & Bowers, 1994). A more comprehensive comparison of cheating trends (Whitley, 1998) used cheating estimates from 86 studies and/or surveys conducted between late 1960s and the mid-1990s. Cheating on exams increased from approximately 25% in years 1976-80, to 32% in 1981-85, and finally to 47% in 1991-95. However, rates obtained in 1969-75 were also quite high at approximately 45%. Not only do first-year college students report cheating less often than high school students, upper-class students (juniors and seniors) cheat less than underclass students which may reflect changes in the college student population over that time frame (Miller, Murdock, Anderman, & Poindexter, 2007).
How Do Students Cheat?
The most studied and prevalent types of academic dishonesty include cheating on examinations, gaining unauthorized assistance on out-of-class assignments, and plagiarizing the work of others as one’s own. Although students have developed a myriad of ways to be academically dishonest, the most common methods of cheating on examinations continue to be strategies that have been used for generations. When taking examinations in classroom setting, the use of pre-prepared crib sheets and copying off others’ papers rank as the most frequent methods of dishonesty among middle high school and college students (Garavalia et al., 2007). Data on younger students suggest they limit themselves to copying off neighbors. As seen in the Prime Time special on classroom cheating (Velmans & Koppel, 2004), crib sheets appear in many creative places such as under the brim of a hat, inside the label of water bottle facing inward even taped to the writing instrument. Less frequently, high school and college students get copies of the exam ahead of time by taking it from the instructor’s desk or computer or from files of tests kept by former students in the same class. Sharing information with peers who will take the same test at a later point during the day or week is also common, but more teachers than students see this behavior as cheating. Similarly, some students provide fake excuses for missing an exam, either to allow themselves more time to study or to gain information about the exam from people who have already taken it (Garavalia et al., 2007).
Technology has afforded students of this generation the ability to instantly communicate with one another, stay in constant touch with world events, and tap into most forms of media “on demand.” This technological explosion has concomitantly created more ways for students to cheat on in class examinations as well as on out-of-class assignments. High school and college students both report that one of the things that make technological cheating easier is their superior facility with technology as compared to their teachers.
Cheating on in-class examination is facilitated by portable electronic devices such as cellular telephones, PDAs, MP3 players, iPods, and programmable calculators. Students use the camera function in their cell phones or other devices to snap pictures of the examination that they share with others who will take the exam later. They can also send those photos with the filled-in answers to students who are taking the exam in the same room, as well as text message responses to specific items on the test. Portable music and video devices are so small that students can load information on them and retrieve it during the exam without alerting the instructor to the behavior. Classrooms are routinely set up for wireless Internet access in most universities, and increasingly in K-12 classrooms as well. A phone or PDA with Internet access can be used to communicate with people outside the exam setting and to pull information off the Internet, including students’ own personal Web pages where they may have stored their notes for a particular class. Even very inexpensive calculators have some kind of memory which students can use not only to store equations, but any other information that they want to have readily available (Velmans & Koppel, 2004).
Online quizzes and examinations, which are increasingly a part of students’ college experience, create their own potential for abuse. Most online exam delivery systems have features such as item randomization, time limits, and item-by-item presentation that can help reduce the potential of cheating. However, unless the computer is equipped with hardware and/or software that verifies who is taking the exam, it is impossible to determine who is actually taking it.
Outside of class, students cheat by sharing answers to homework assignments and plagiarizing papers they hand in as their own. The World Wide Web has abundant resources to help students complete these types of assignments ranging from services that provide complete papers to an array of Web sites that have well-written syntheses of information such as online encyclopedias like Wikipedia or the innumerable Web sites posted by course instructors for their classes. Google (google.com) can be used to identify papers on a given topic, or scholars in a given area, often with links to their works. Sites that explicitly sell papers to students on the topic of their choice abound, with prices determined by a combination of desired quality, length, and the amount of customization that is required.
Despite the volumes of published research articles on academic cheating, few theoretical models have emerged to frame this body of scholarship. However, findings from a series of pivotal and now classic studies of Hartshorne and May (1928) clearly suggest that any such model must incorporate both individual and environmental elements, as well as the interaction among them. These scholars initially set out to determine the character traits that were associated with people who made more or less moral decisions across a variety of situations, but found little cross-situational consistency. Although many of the thousands of students they studied cheated on their school work some of the time, few students cheated in every assessment situation and making the decision to be honest in the classroom was not always consistent with their behavior in other domains. Recent efforts to identify the profile of the moral student, typically account for a small amount of the variance in students’ actual behavior (Murdock & Stephens, 2007). As such, the view of dishonesty as a person by environment interaction undergirds most recent theoretical models that have attempting to provide a framework for understanding cheating.
Moral Development and Social Cognition Applied to Cheating
Moral development scholars have applied Kohlberg’s (1984) stages of moral reasoning to cheating issues. From this perspective, students are seen as more morally developed when they reason about moral issues based on abstract universal moral principles (Kohlberg postconventional reasoning, stages 5 and 6) rather than relying on external cues from the environment, such as rewards and punishments (Kohlberg preconventional reasoning, stages 1 and 2) or on definitions of what is right and wrong as prescribed by religion, laws, or other forms of authority (Kohlberg conventional reasoning, stages 3 and 4). Although students report that their personal morality is one of the main reasons that they do not cheat. Blasi (1980) found support for this hypothesized cognition-behavior connection in only 7 of 17 experimental studies. Relationships between moral reasoning and behavior in correlation studies are only moderate (Whitely, 1998). More recent experimental studies also find that students with higher levels of morality do not consistently exhibit less dishonest behavior than those with lower levels of morality, and the effects of moral reasoning may depend on the context (Murdock & Stephens, 2007). For example, when participants are being watched by the experimenter, termed high surveillance (Corcoran & Rotter, 1987) or are given strong incentives to do well (Malinowski & Smith, 1985) those with stronger higher morality scores cheat less than those with lower scores, but no differences between higher and lower morality groups were found in either the low-surveillance or low-incentive conditions.
One explanation for the lack of clarity in the relations between moral reasoning and moral behavior comes from Turiel’s (1983) domain theory of social reasoning which suggests that decisions about cheating are not viewed by most people as a moral issue, but rather as an issue of social convention. Thus, contextualized attitudes about the behavior rather than more global moral principles should govern cheating decisions. Indeed, these more contextualized cheating attitudes are consistently among the strongest predictors of dishonesty in two published reviews of cheating literature (Bushway & Nash, 1977; Whitley, 1998). These cheating attitudes are often referred to as “neutralizing attitudes,” a term coined by Sykes and Matza (1957) to capture how people deflect responsibility for deviant behavior. Students mainly neutralize the decision to cheat by denying that it is a crime, pointing out that others are not harmed by it, or by shifting the responsibility for the behavior away from themselves by blaming cultural norms, the competitive context, or poor teaching, among other things (Murdock & Stephens, 2007). Thus, whereas the moral reasoning perspective focuses only on the characteristics of the individual, a social reasoning perspective suggests that attitudes toward cheating are a function of people in context.
Theory of Planned Behavior Applied to Cheating
Whitely (1998) developed a model of cheating behavior grounded in Azjen’s theory of planned behavior as applied to dishonest action (Beck & Azjen, 1981). According to this model, people’s intentions to cheat are the main determinant of their actual behavior. These intentions develop as a result of person-environment interactions that contribute to students’ (1) moral decision making, (2) perceived norms related to cheating, and (3) attitudes about the acceptability of cheating. Students who feel less of a moral obligation not to cheat, who perceive cheating as more normative, and/or who see cheating as more justifiable or easily neutralized as “okay” will have stronger intentions to cheat.
Two factors influencing students’ intentions to cheat are the perceived benefits of cheating and the perceived risks associated with being caught. According to this model, students who will benefit most from cheating are those who have high needs for approval, who frame success in terms of performance and grades versus learning, and who have reasons to doubt their own ability to accomplish their desired level of performance. Both individual factors such as perfectionism and prior achievement and contextual factors such as parental pressure, time available for studying and classroom competitiveness determine the expected benefits from cheating. Similarly, the perceived risks of cheating are a function of one’s assessed cheating skill coupled with the conditions in an individual classroom that make cheating easier or more difficult as well as the punishment associated with being caught. Situational factors which affect the ease with which cheating can be completed without detection also moderate the relationship between intentions and behavior: Given two students with equally strong intentions to cheat, the student who is in the situation with more perceived situational risk of being caught will be less likely to actually cheat.
Achievement Motivation Approaches to Cheating
Murdock and Anderman (2006) applied constructs from theories of academic achievement motivation to frame the research on academic cheating. They propose that students’ decisions to cheat are influenced by the individual and contextual variables that affect three questions: What do I hope to accomplish? Can I accomplish what I want? and What are the costs associated with cheating?
Students whose academic goals include intrinsic enjoyment, interest or mastery of material for some future goal are posited as less apt to cheat that those who are simply trying to attain some particular grade, or who are simply trying to get by. Personal factors influencing these goals include students’ own interests, their theories of intelligence, and need for accomplishment as defined by grades or other external rewards; classroom pedagogical practices, parental pressures, and school policies such as grade point average requirements for athletic eligibility presumably contribute to these goals.
Clearly, not all students whose goals are straight As or who need to achieve a 2.0 average to play soccer are going to cheat. Those who believe they have the skills needed to accomplish those goals will cheat less frequently that those with more doubts about their ability. Doubts about the ability to accomplish one’s goals might be higher for students with poorer histories of past achievement relative to one’s goals, inadequate study strategies or study time, or high levels of anxiety during testing situations. Classroom practices also influence students’ expectations. For example, teachers whose pedagogy is not clear, whose objectives, assessment, and instruction do not align, or who adopt any other behavior such as norm-referenced grading that is perceived by students as adversely affecting their ability to achieve their goals are lowering students’ expectations for success, and may inadvertently increase cheating. Finally, students are anticipated to cheat less often when the costs of cheating are higher. In addition to the costs of detection as described in Whitley’s (1998) model, these costs include the cost to one’s self-image or violating one’s personal values or beliefs about the acceptability of cheating.
Theoretical Models: Summary
What emerges as consistent across all of the contemporary perspectives on cheating is that students make decisions to cheat within contexts that provide both incentives and disincentives for being dishonest. Further, these incentives vary with the situation. Students approach each learning situation with their own personal histories, attitudes, and motives, which undoubtedly shape both what they see and how they chose to respond to their constructed realities. What are the major individual and contextual variables that have been found to predict dishonesty?
Individual predictors of dishonesty. There has been tremendous interest in trying to determine what makes someone a cheater. Although the available evidence shows that few people consistently cheat and that environmental factors play a large role in determining when cheating occurs, there are some personal attributes that appear to be linked to higher levels of academic dishonesty.
At the college level, Whitley (1998) found that more frequent cheaters have different beliefs about cheating than their less dishonest peers as measured by more positive attitudes about cheating and less moral obligation not to cheat. At a more general level, increased rates of cheating were associated with students seeing themselves as less honest and as well as having engaged in more prior cheating and other deviant activities. Finally, dishonesty also appears to increase among students whose study skills are more lacking and whose perceived pressure for successes is higher. These factors are much more closely related to cheating than are actual level of prior achievement, which is only minimally related to dishonesty.
Although no comprehensive literature review exists at the K-12 level, several individual variables repeatedly emerge as related to cheating. Motivationally, students with lower levels of self-efficacy, whose reasons for academic engagement are less about learning and more about grades, and who view intelligence as something that is fixed rather than malleable seem to cheat more frequently than their peers (Miller et al., 2007). Similarly to college students, middle and high schools students who cheat hold more permissive attitudes about cheating than their more academically honest classmates.
Few demographic factors are consistently related to cheating at any level of education other than the age and grade relationships. At the K-12 level, age and grade in school are linearly associated with increased cheating, whereas among the university population, younger students and nonmarried students cheat more often than those who are not married. It may appear that males cheat more often than females, but a closer look at the research suggests that these differences may be largely an artifact of how cheating is defined. Male students typically have higher rates of self-reported dishonesty and report more tolerant attitudes toward cheating than female students. Yet, some studies disconfirm these findings and others suggest that the difference may be largely due to the way the data on cheating were collected. For example, when gender comparisons are based on actual rather than self-reported behavior, differences between males and females are minimal, suggesting females may be more motivated to appear honest. High school females also admit to cheating as often as males when asked about cheating in the service of helping others, again suggesting that gender roles and social desirability may influence self-reports of dishonesty (Miller et al., 2007).
Contextual Predictors of Academic Dishonesty
The prevalence of dishonest behavior appears to be influenced by students’ peers, classroom policies, and factors outside of school.
Peer norms. Rates of cheating are higher among students who view their peers as accepting of academic dishonesty also are more likely to cheat (Whitely, 1998). However, although students who cheat indicate that more of their peers are cheaters, it is not known if there are actual differences in the behavior of the peer groups or simply perceived differences.
Classroom context. Pedagogical and interpersonal aspects of the classroom appear to influence both actual rates of cheating (Whitely, 1998) and the extent to which students report that cheating is justifiable (Murdock & Stephens, 2007). Pressure for grades repeatedly emerges as contributing to cheating and several aspects of the classroom can contribute to this. For example, there is some evidence that students cheat less often in classrooms that are characterized by having a mastery goal structure or a strong emphasis on learning and improvement versus a predefined level of accomplishment (Anderman, 2007) as well as in cases when teachers grade on a competitive curve. In addition, cheating decreases when students have a genuine interest in a given task as opposed to simply working for the grade (Schraw et al., 2007).
Rates of student honesty may also be linked to teachers’ demonstrations of what Wentzel (1997) refers to as pedagogical caring, which is characterized by instructional soundness and interpersonal caring. From middle school forward, students indicate they cheat less when teachers have clear instructional objectives and are prepared and able to facilitate learning toward those objectives. Interper-sonally, pedagogical soundness includes treating students with respect and impartiality. Finally, students indicate that one of the main factors contributing to their cheating is having tests that are not fair; that is, those where the material is not consistent with what was taught or are overly lengthy or difficult (Whitley & Keith-Spiegel, 2002). However, although poor instruction, unfair testing, and perceived disrespect are known to justify students’ cheating (Murdock, Miller, & Goetzinger, 2007) and it seems logical from many theoretical perspectives that these behaviors would lead to increased cheating, the extent to which they actually cause cheating is still unclear.
Outside context. Pressure for grades and lack of time available to study can also be influenced by other factors in students lives including family expectations, extracurricular involvement, full- or part-time jobs, and family responsibilities (Whitely, 1998).
Over 90% of the empirical studies on academic cheating assess dishonesty based on students’ self-reported behavior. This methodology has several limitations. First, we know that students are not always honest in reporting about their behavior. In the 2006 Josephson Survey of Youth Ethics, 27% of the adolescent respondents admitted to have lied at least one time on the survey. In short, they lie about lying! In addition, the definitional issues that are inherent in delimiting what is meant by cheating affect researchers as well. Some studies simply ask about cheating, whereas others specify particular behaviors ranging from those that many would not see as cheating, such as sharing homework, to others on which there is a general consensus. Of the large number of studies that look at predictors of cheating, some limit cheating incidents to the past year, others to a single class and others to a still smaller time period and context. Having such a large range of times and contexts in which the behavior in question is being measured makes accurate and consistent estimates of predictors quite difficult.
Methods that rely on behavioral indicators of cheating include both experimental laboratory studies and, less frequently, classroom-based research. Participants in laboratory studies are often given a task that they could only succeed at by cheating, such as tracing a complex series of shapes with one’s eyes closed, or they are left alone to complete a task with the answer key for the task in full view of the participants. Although the measurement of the dependent variable may be more accurate in these cases than in self-report studies, given the known influence of environmental variables of cheating behavior, there are questions about the extent to which results from these laboratory studies would generalize to actual academic settings.
A third group of studies measures cheating behavior in actual classroom or testing situations. One common method uses software programs based on probability theory to identity likely cheaters based on the unlikelihood of the student having given a specific set of answers or by estimating their expected and actual concordance between their answers and those seated around them (Haney & Clarke, 2007). Other scholars have provided students with opportunities to cheat by allowing them to grade their own exams after having secretly photocopied the original tests, or making copies of phony answer keys accessible but off-limits. These methods have more ecological validity than laboratory tasks and are less subject to response bias than self-reports; the limitation is that they only assess cheating via a particular method in a particular context.
Prevention of Cheating in the Classroom
Although most research on cheating prevention has been conducted in higher education settings, there is no reason to assume the suggestions that arise from this research would not generalize across settings. One important and easy to implement strategy is to set clear expectations for and definitions of what constitutes honest behavior, delineate consequences for violating those expectations, and be willing to take the steps necessary to enforce those consequences (McCabe, Trevino, & Butterfield, 1999). In short, a commitment to integrity has to be modeled by the institution, whatever the educational level (Whitely & Keith-Speigel, 2001). Although a growing body of research suggests that institutionalizing these expectations into formal honor codes may increase students’ honesty in the classroom, McCabe and Trevino (1993) found that it was the institutional commitment not the code itself that matters.
Providing students with clear objectives and with instruction and assessments that match those objectives is the another basic component in promoting honesty. When students sit down to take a quiz or examination, the content, format, and level of difficulty should not surprise them. Paper assignments that are to be completed outside of class should have clear, specific directions and scoring rubrics, and appropriate levels of scaffolding available so that students can be successful if they wish to. The more frequently that students are assessed, the more likely they are to stay caught up with the information, making cramming less likely, and the less each assessment counts toward the overall grade, so that “high-stakes” situations are avoided.
Cheating can be lessened by promoting students’ investment in their learning. Given that students are less apt to cheat when they are personally interested in what they learn (Schraw et al., 2007) and are motivated toward learning rather than simply achieving a given grade (Anderman, 2007), providing students with some input into what they learn, setting tasks at the appropriate level of challenge, and creating an environment that encourages curiosity, risk taking, and improvement should all promote honest behavior.
Exams. Exams can be formatted to make it more difficult to cheat. For example, the most effective way to reduce cheating on forced-choice examinations is to develop multiple versions of the test in which both the answer choices as well as the item stem itself are randomized across versions. Students cheat just as efficiently on tests when the questions are scattered than they do if everyone has the same version (Houston, 1983). Examinations should not be easily identifiable as being a particular version, as is the case if different colors of paper were used. In most middle and high school classrooms, there is enough space to spread out, so that two copies of an exam should suffice. In many college classrooms, students are tightly spaced and are as likely to copy off someone who sits in front of them as they are someone next to them. These situations require three or four different test versions. Assigning students to seats during exams also eliminates their ability to set themselves next to someone who they may have prepared to cheat with (Houston, 1986).
When exams are essay, short answer, or problem based, the order of questioning should still be changed and students should be provided with a piece of blank paper and asked to cover their responses as they write. In some cases, parallel exams can be easily developed by leaving the problems intact but changing certain details such as the numbers. With all examinations, it is important that the instructor make sure they are not accessible to students before they take the test. This may mean creating multiple test versions if the same course is taught at different times of the day or week, keeping exams in locked paper or electronic files, and not using the same exams year after year (Cizek, 1999).
Although some instructors believe that a laissez-faire approach during the exam itself communicates trust to students and thereby increases honesty, current data do not support this. Students cheat less when the instructor appears alert and attentive during an exam, circulates around the classroom, and is aware and responds to suspicious behavior. Many students see these behaviors as the instructor’s job and become frustrated when blatant cheating occurs under the nose of an oblivious teacher (Cizek, 1999).
Plagiarism. Plagiarism can be reduced at a multitude of points in the process. Most of these strategies are simply good pedagogy: they not only decrease cheating, but they also increase learning. When designing writing tasks, consider steering away from assignments that are likely to be given by hundreds of instructors across the United States, such a biography of a U.S. president. Instead, develop an assignment that requires the student to deal with the desired information in a particular context or develop a particular opinion. At any grade level, it is often useful for student to hand in pieces of a large project along the way for feedback. These checkpoints reduce the anxiety that may be created in students who cannot plan and therefore try to cram before the due date. They give students more control over their final grade because the feedback gives them control over their learning. Finally, feedback on drafts can include reports generated from commercially available plagiarism detection programs such as http://Turnitin.com so that student learns as they go when they are relying too closely on the work of others. It is equally important that students are taught what it means to plagiarize.
The substantial empirical research base on cheating is dominated by efforts to predict who will and who will not cheat and under what circumstances. Yet, several important questions remain understudied including the effects of accountability movements on academic integrity, pedagogical practices that will encourage honesty, and the ways in which institutions can support teachers and professors as they deal with dishonesty.
Accountability Movements and Cheating
Recent reform movements encourage large-scale testing to ensure that school districts are teaching students to a defined standard. Termed accountability, there is beginning to be evidence that these high-stakes tests are creating situations that encourage teachers and other administrative personnel to cheat. Nichols and Berliner (2007) point out, however, that schools are also being asked to create the same outputs even though they are working with dissimilar human and economic capital. In a classroom context, these are conditions that students would consider ripe for justifying cheating, and more thorough attention should be given to how policies that foster competition for resources affect classroom behavior (Callahan, 2005).
Pedagogical Practices and Cheating
Teaching that is characterized by sound pedagogical practices and engaging curriculum appears to foster a motivation to learn and reduces cheating. Despite the numerous studies that make these claims, the results are inconclusive largely because they simply relate a given student’s report of what the teacher does in the classroom to his or her own behavior. Needed are studies that actually evaluate the effects of improving instruction by providing students with more choices, focusing curriculum on interesting real-world problems, carefully aligning instruction and assessment and other factors that have been linked to reduced dishonesty.
Attention to the Needs of Teachers and Faculty
Cheating is seen as a serious problem by teachers and school personnel across all levels of education, yet we also know that many teachers are reluctant to confront these issues in their own classrooms (Simon et al., 2003; Spiegel, Tabachnick, Whitley, & Washburn, 1998). This reticence appears to be affected by a myriad of factors that educational institutions could help to ameliorate including perceptions of poor support from upper administration, fear of retaliation, and the personal anxiety of dealing with the situation. Future studies might focus on identifying specific practices that could support teachers in these areas so that appropriate interventions might be designed and evaluated.