When Opportunity Knocks: College Students’ Cheating Amid the COVID-19 Pandemic

Abstract

Background

The COVID-19 pandemic led to significant, unforeseen changes in classroom instructions, including the evaluation of students.

Objective

The purpose of this study was to investigate college students’ cheating both before and during the COVID-19 pandemic in terms of (a) preponderance of cheating, (b) the factors that may have led to an increase in the amount of cheating, and (c) the underlying reasons for and affective response to cheating.

Method

A sample of primarily Psychology majors (N = 214) attending a public land-grant university in the southeastern U.S. voluntarily completed a survey at the end of the Fall 2020 semester.

Results

The results showed that the COVID-19 pandemic increased first time cheating, cheating in online classes was higher than that of in-person classes for most types of graded materials, and students are adept and adaptive at dealing with faculty attempts to combat cheating. Students’ primary reasons for cheating were “feeling pressure,” and “pandemic,” and students who had cheated reported feeling “relieved” most often.

Conclusion

With the onset of the pandemic and subsequent increase in online instruction, cheating behavior has also increased.

Teaching Implications

As online enrollment continues to grow, understanding students’ cheating behavior.

Keywords

higher education COVID-19 cheating online education undergraduate psychology

While cheating in college has been an issue for as long as universities have existed, cheating behavior has steadily increased as time has gone on. With more and more universities offering remote learning, college cheating has only become more rampant. Now, as Coronavirus Disease 2019 (COVID-19) has had global health and financial impacts, it has also had an impact on the academic integrity of college students (Gamage et al., 2020). It is important to define cheating in the current context, as previous research has differed in their definitions (see McCabe et al., 2012). Cheating behavior, as we will use it, occurs when students wrongfully use any information of which they are not the intellectual owner, on academic assignments that are considered part of their final grade (University of Kentucky, 2020). In thinking about cheating, it is important to note that cheating in college can occur during in-person classes or online classes, and that cheating can include: (a) offline resources—anything where one exchanges or obtains information NOT using an electronic device. and (b) online resources—anything where one exchanges or obtains information using an electronic device (including texting). The purpose of the present study was to investigate college students’ cheating amid the COVID-19 pandemic in terms of preponderance of cheating, the factors that may have led to an increase in the amount of cheating (e.g., more online courses), and the underlying reasons for and affective response to cheating.

Students cheating in college is a problem that has plagued the Academy since the inception of the university system. Academic honor codes against cheating date back to the early 1700s at The College of William and Mary (Danilyuk, 2019). Like today’s honor codes at all types and levels of schools, the William and Mary Honor Code had students make a verbal pledge to never lie, cheat, or steal in both their academic career and outside of school. Given the long history of cheating, it is not surprising that research has shown an overall high level of this behavior. This research includes the classic study of Bowers (1964) that included a multi-campus study of almost 5000 students from 99 institutions. In his research, Bowers examined 13 different types of cheating (e.g., plagiarizing work, copying answers, or using notes during an exam) and asked students if they committed each act. He found that 75% of students admitted to at least one type of cheating and that more than half admitted to two. Later research supported Bowers (1964). In particular, McCabe (2005, 2020; McCabe et al., 2012) presented data from surveys of about 70,000 undergraduates (excluding first-year students) between Fall 2002 and Spring 2015. McCabe used a subset of nine of the 13 Bowers questions. The results showed that 39% admitted to cheating on tests and exams, 62% admitted to cheating on written assignments, and 68% admitted to cheating on both tests/exams and written assignments. In addition to McCabe’s large data sets, there are many examples of cheating scandals at schools throughout the US. This includes elite colleges such as Harvard in 2012 (Carmichael, 2012), Stanford in 2015 (Seipel, 2015), and Yale in 2019 (Prihar & Wanna, 2019).

It is unclear if there is one motivating factor as to why college cheating takes place, often, there are likely several contributing reasons (i.e., peer pressure, ambiguous expectations both at the instructor and university level, pressure to succeed, etc.; McCabe et al., 2012). However, cheating habits probably develop long before students get to college, particularly during one’s high school years (McCabe et al., 2012). Moreover, researchers have examined the causes of cheating from various perspectives. First, some researchers have investigated individual differences. For instance, Harbin and Humphrey (2013) note that level of morality may guide students’ cheating (see Kohlberg, 1971). Second, demographic factors have been implicated in cheating. Yu et al. (2017) replicated prior research findings (see Whitley et al., 1999) that female students reported cheating less than male students. It is also interesting to note that Yu et al. found a positive association between student involvement in extracurricular activities (including working) and cheating (see also McCabe et al., 2012). This might reflect that active students are unable to spend time studying and thus resort to cheating. Third, some research has focused on how pressure may impact a student’s willingness to cheat, including pressure to succeed, peer pressure, or pressure due to a heavy workload (see Danilyuk, 2019; Miller et al., 2017).

Fourth, cheating has been conceptualized as a decision-making process. Murdock and Anderman (2006) argued that students ask themselves three questions when deciding to cheat: what the purpose of cheating is, if they can successfully complete a task, and what the cost of cheating is. To illustrate, if the student determines that cheating will lead to a good grade, then the purpose of cheating is to earn better grades. If the student decides that they cannot successfully complete a task, they will likely cheat to finish it. Regarding the potential costs of cheating, a student understands that any decision to cheat has the possibility of getting caught and being penalized. One must keep in mind that punishment includes specific academic consequences (e.g., loss of points, failure in the course, and suspension) as well as a loss to one’s self-image (i.e., integrity). If the likelihood of getting caught is relatively low, cheating is likely to occur.

A final reason why cheating may occur, related to decision-making, is that cheating occurs as a function of opportunity (Adzima, 2020; Harbin & Humphrey, 2013; Hylton et al., 2016). The issue of opportunity has recently become a point of concern with the proliferation of online education in higher education (see Young, 2012). Keep in mind that online education within higher education in the United States has skyrocketed in recent years, increasing from 15% of college students enrolled exclusively in online courses during the Fall 2017 semester to 97% of students in June 2020 due to COVID-19 (Bustamante, 2020). Researchers (e.g., Harbin & Humphrey, 2013; Hylton et al., 2016; Varvel, 2005) have argued that online education increases the opportunity for cheating. This is primarily because online education leads to relative anonymity and separation between instructor and student. Not only can an instructor not directly observe a student taking an exam in person, but the student typically has unfettered access to information from the Internet or from other resources. We should add that recent technological advances that allow instructors to monitor students are not fool proof. An instructor can lock a student’s ability to use the Internet on the computer an exam is taken, but a student can simply have another computer or their cellphone available to search the Internet. One might argue that the preceding problem can be addressed by having an instructor use video-monitoring software to observe a student’s behavior during an exam. However, this software may lead to the instructor having to view many videos (especially for large survey courses), and such observation cannot definitively indicate that cheating occurred. That is, a student looking at their cellphone during an exam may be the result of a student reading a text from their parent rather than an attempt to cheat. While studies show that cheating does occur with online classes (see Young, 2012), the results of surveys comparing in-person and online cheating are ambiguous. Some studies have found no difference (Grijalva et al., 2006; Watson & Sottile, 2010), but others show an increase of cheating in online classes compared to in-person classes (e.g., Lanier, 2006). Another study found that students in online courses and in-person courses have varying ideas about what actions are considered cheating behaviors (Burgason et al., 2019). For example, more in-person students felt that using notes was considered cheating compared to online students.

Regarding students cheating in online courses, if students feel anonymous and unlikely to be adequately monitored, they may assume that the likelihood of being caught cheating is virtually zero and cheat more in online classes using online resources (e.g., the Internet). Previous research has shown that participants had a higher propensity to cheat when chances of being caught were less likely (Kajackaite & Gneezy, 2017). Kajackaite and Gneezy had participants play one of two versions of a basic mind game in which students were asked to either privately think of a number before rolling a die or roll a number five. The primary difference between the two games was that participants who privately thought of a number had zero risk of being caught cheating, while participants who had to roll a number five could lie but would likely be caught. If participants reported that the number the die landed on matched the number they had thought of or the number five, they were rewarded with varying amounts of money ($1, $5, $20, and $50). As the reward money increased in value, so did participants’ willingness to cheat, but only when their perceived risk of being caught cheating was zero. In other words, when participants assessed the benefit of cheating to outweigh the cost, they were more likely to cheat.

Since the beginning of the COVID-19 pandemic, concerns about cheating in online classes have increased. As the number of COVID-19 cases rapidly grew in the United States, many colleges had to shut down and switch to primarily online instruction. This reformatting was rapid for most colleges, as 98% of college institutions had switched from an in-person to online format (Bustamente, 2020). Both instructors and students had to adapt to this new pedagogical approach, most who had minimal experience with online education. This transition was difficult for both parties for several reasons. Instructors had to learn to teach remotely using new technology they likely had little prior experience with (Feder, 2020). While technological advances such as Zoom and Microsoft Teams has helped the transition greatly, this meant that instructors had to take on learning the new technology while also dealing with the stress that COVID-19 naturally brought about (worrying about one’s own health, family health, job security for non-tenured professors, etc.). Students, however, likely faced other difficulties during this transitional period. Reliable internet access at home, access to technology, and even differing time zones are all examples of the challenges students had to face. Institutions additionally grappled with the challenge of ensuring academic integrity while attempting to provide equitable resources for students (Feder, 2020).

Recent research has helped to reveal the connection between COVID-19 and the uptick in cheating behavior. Specifically, Gamage et al. (2020) highlight several reasons for the increase in cheating. Namely, restrictions for assignments are harder to enforce via distance learning, students had less educational support (i.e., library resources) during the pandemic, and the challenge of identifying academic misconduct remotely. As a result, the pandemic has seen cheating scandals related to online cheating at various institutions (e.g., Georgia Tech and Boston University, Miller, 2020; the U.S. Military Academy at West Point, CBS News, 2020; Texas A&M, McGee, 2020; Subin, 2021).

In addition, it is unclear what the specific nature of the cheating is (e.g., in online classes using online resources) as well as what the underlying reasons motivating students to cheat in an online learning environment are. Regarding the latter, it is unclear whether an increase in cheating is due to greater opportunities for cheating (e.g., group messaging applications such as “GroupMe” or an educational website such as “Chegg”) or whether other factors related to the COVID-19 pandemic are implicated (such as time-management issues; see Harbin & Humphrey, 2013, issues with technology, or lack of resources needed for academic achievement; Feder, 2020). Finally, if cheating in an online learning environment has increased, how do students feel (i.e., their affective response) about cheating (e.g., feeling guilty for cheating vs. not caring about this behavior).

In the present study, we administered a self-designed survey about cheating to undergraduates at a Research I university. The focus of the survey was primary on type of graded material (exams, quizzes, homework assignments, and projects/papers), the class format (online and in-person), and type of cheating resources students used (online or offline resources). Participants were asked to report whether they had cheated in differing instances, such as in an in-person class on an exam using their notes. We also focused on the COVID-19 pandemic, asking participants if they had cheated for the first before or after the start of the COVID-19 pandemic and subsequent switch from an in-person to an online format. To gain better insight, we also asked participants open-ended questions about various general aspects of cheating (e.g., Why did you cheat? How did you feel if you cheated?) and more specific questions about their experience with software that helps to protect against online cheating. The inclusion of these open-ended questions allowed us to understand student cheating in a unique way, as most previous research investigating online cheating has not included such questions. If these questions were used, responses from a small subset of students were usually presented as illustrating a particular point (see McCabe et al., 2012), the full data set is not analyzed.

Based on prior research, we had the following five hypotheses:

Hypothesis 1: Overall Cheating

We predicted that participants would acknowledge cheating overall and across all types of graded materials (exams, quizzes, homework, and project/paper; see McCabe, 2020). Moreover, it was predicted that very few students would report being caught cheating (Prihar & Wanna, 2019).

Hypothesis 2

First Time Cheating after COVID-19

We predicted that the COVID-19 pandemic would provide greater opportunities for students to cheat for the first time. Specifically, we expected that the switch to online classes would provide greater opportunities for students to cheating using online resources (e.g., Google, Group-Me, etc.) and that cheating would be higher for exams, quizzes, and homework than for projects/papers. As the pandemic likely impacted the mental health of college students (see Usher et al., 2021), it is possible that students were more likely to cheat to avoid further stress.

Hypothesis 3: Use of Software to Prevent Cheating Using Online Resources

We predicted that based on the increased use of software to prevent cheating using online resources during the pandemic (Harwell, 2020), students would acknowledge that their instructors used this software and would report trying to circumvent this software.

Hypothesis 4: Impact of Demographic Variables

It was anticipated that demographic variables would lead to main effects on one’s decision to cheat. For example, there would be more cheating by underclassmen than upperclassmen (McCabe et al., 2001), the highest degree of cheating would be done by students with the lowest grade point average (McCabe et al., 2012), and students who are more engaged at school (e.g., Greek life, athletics) or employed would report cheating overall more than students without a job (Danilyuk, 2019; McCabe et al., 2012).

Exploratory Analyses

A series of exploratory analyses were conducted to investigate various aspects of cheating for which there has been relatively little prior research (i.e., no clear hypotheses could be generated). First, based on Harbin and Humphrey (2013) view of cheating and opportunity, it is likely that for online classes there is a greater opportunity to cheat (using both online and offline resources) on graded material such as exams, quizzes, and homework using online resources. This opportunity is the result of students being in a physically different location than an instructor and thus all material to take an exam/quiz or complete homework being readily available. Regarding a project/paper, it is unlikely that students would report cheating more often on online courses than in-person courses because this type of graded work is not time constrained (i.e., it is due at the end of a semester) and there is less concern with generating a specific accurate answer. For in-person classes, the pattern of results is likely to be very different. The greatest opportunity to cheat in these classes might be on homework since it is completed out of sight of the instructor and requires a student to answer accurately in a relatively short amount of time. Moreover, as with online classes, cheating on homework can be accomplished out of the classroom using online and offline resources. Conversely, the amount of cheating on the other types of graded material would likely be reduced because: (a) for exams/quizzes cheating in an in-person class is relatively difficult (using either online or offline resources) to carry out in the presence of an instructor, and (b) for project/paper the same issues described above for online classes (i.e., lack of time constraints and less concern with generating an accurate answer) is still true. We should add that the above patterns should lead to differences in the amount of cheating between online and in-person classes for exams and quizzes, but not for homework and projects/papers.

Second, we examined frequencies of responses to Yes-No questions (e.g., would you turn someone in who cheated). It may be the case that very few students will admit to turning in another student for cheating. Finally, some open-ended questions were analyzed by examining participants’ cognitive representations of why they cheated and how they felt after cheating using networks of the participants’ mental models. The models were created using Pathfinder analysis, a psychometrically established scaling tool (Schvaneveldt, 1990). Pathfinder analysis creates a representation of the data based on the similarity between concepts in semantic memory.

Method

Participants

Participants were recruited via email invitation. A total of 1748 students with Psychology as their declared major or minor were contacted to participate. Overall, the response rate was 12.24%. The consenting participants (N = 214) were undergraduates at a large Southeastern university. Participants’ age ranged from 18 to 67 years (M = 21.63, SD = 6.34). The mean number of credit hours completed by students was 78.82 (SD = 38.63, range 0–160)—students need to complete 120 credit hours to graduate at this university. Students reported a cumulative grade point average of 3.54 (SD = 0.46, range 1.95–4.00). Further participant demographic information is available in Table 1.

Table 1.

Participant demographics

	n	%
Gender
Women	181	84.6
Men	29	13.6
Transgender	1	0.5
Other	3	1.4
Race
White	163	76.2
Black/African American	19	8.9
Hispanic/Latinx	10	4.7
Asian/Pacific Islander	15	7.0
Other	7	3.3
College year
Freshman	29	14.7
Sophomores	34	17.2
Juniors	44	23.3
Seniors	90	45.7
No answer	17	0.08
College career
Psychology major	126	58.9
Psychology double major	38	17.8
Psychology minor	27	12.7
Other major	13	.06
Involved in Greek life
N	56	26.2
Employed
N	131	61.2

Design

The current study consisted of a self-created survey. The survey asked several questions about cheating—type of material (exam, quiz, homework, and project/paper), class format (online, in-person), and type of cheating resource (online, offline). In total, the survey consisted of 51 questions, including demographic questions and follow-up questions (of which were only presented if a student had marked ‘yes’ for some questions and ‘no’ for others). The primary dependent variables of interest were whether participants admitted to cheating to any of the types of material, if they had cheated for first time before COVID-19, what type of resources students used to cheat, and participants’ affective open-ended responses to why they decided to cheat and how cheating had made them feel.

Measures

The survey included three parts (see Jenkins et al., 2021 for survey materials). First, questions were initially asked about participants’ age, gender, ethnicity, major, minor, number of credit hours completed, overall grade point average (GPA), membership in a sorority or fraternity, and working status.

Prior to the second part of the survey, participants were presented with a definition of cheating (University of Kentucky, 2020):

Cheating is defined by its general usage. It includes, but is not limited to, the wrongfully giving, taking, or presenting any information or material by a student with the intent of aiding himself/herself or another on any academic work which is considered in any way in the determination of the final grade.

Next, students were informed that in answering the survey, they were to consider two types of college classes:

a. Online—any college class that is taught partially or fully online, in an asynchronous or synchronous manner.

b. In-person—any college class that is taught face-to-face in a room on campus

Finally, students read that cheating can include:

a. Online resources—anything where you exchange or obtain information using an electronic device (including texting).

b. Offline resources—anything where you exchange or obtain information NOT using an electronic device.

The second part of the survey included questions about cheating. Unlike prior research (e.g., McCabe et al., 2012), we chose not to include specific questions about different types of cheating. As stated earlier, Bowers (1964) and McCabe et al. (2012) used a series of 13 and nine questions, respectively. McCabe included questions about plagiarizing, copying answers on an exam, and turning in another student’s work as their own. Instead, we included 20 questions that pertained to four general graded materials. These included exams and homework as asked by Bowers and McCabe, but also included questions about quizzes and projects/papers. For each of the four graded materials (exams, quizzes, homework, and project/paper), there were five Yes-No questions (Yes = 1, No = 0). Specifically, for each graded material, participants were asked if they cheated in an online class and an in-person class. Upon answering yes, participants were asked if they used online resources and if they used offline resources to cheat. Lastly, if participants answered yes to whether they had cheated, participants were asked a fifth question about whether they had cheated for the first time after COVID-19 pandemic.

The third part of the survey included a combination of 20 questions—10 Yes-No questions and 10 follow-up questions. Specifically, participants were asked Yes-No questions about if they had ever cheated and if they had tried to circumvent anti-cheating software (e.g., Lockdown Browser), to name a few examples. After each Yes-No question, participants were then asked open-ended follow-up questions about how they felt after deciding to or not cheat and how they dealt with anti-cheating software. The open-ended responses for each participant provided descriptive data and were analyzed using Pathfinder analysis. We used Pathfinder analysis to create cognitive representations of participants’ responses. Open-ended questions have been included in prior cheating research, but the answers to these questions have not been used for analysis (e.g., McCabe et al., 2012). In the current study, we analyzed these open-ended questions to determine which concepts were most relevant to participants’ cheating behavior. For instance, it may become clear after analysis that a driving force in participants’ decision to cheat was that they felt overwhelmed by their classwork. Further, COVID-19 may have been a particular motivator for academic dishonesty. While previous research has explored open-ended responses, this data was not used in any form of analyses to understand the thought process behind participants’ decisions and responses.

Procedure

The current study was approved by the university’s Institutional Review Board. Participants were sampled using a voluntary convenience sampling method. Specifically, students majoring and minoring in psychology were invited via LISTSERV email requests to participate in a 15 minute survey about cheating on Qualtrics. The request made clear that although the survey was about cheating, all responses were anonymous and no identifying information was collected. Note that some faculty heard about the study and announced it in their classes—it is unclear how often this occurred. Participation was voluntary and there was no compensation except for students from one introductory psychology course were offered extra credit for participating. The survey was administered during the last 2 weeks of the Fall 2020 semester (i.e., last week in November and first week of December).

Results

Hypothesis 1: Overall Cheating

Hypothesis 1 was supported. The percentage of participants who reported cheating was 74.8% (n = 160) across all types of graded materials (58.4% exams, 61.2% quizzes, 60.7% homework, and 13.6% project/paper). We should note that the high levels of cheating were also reflected in the percentage of participants who cheated on more than one type of graded material. While 20.1% (n = 43) of participants cheated on only one type of material, 22.4% (n = 48) cheated on two types of material, 24.3% (n = 52) cheated on three types of material, and 7.9% (n = 17) cheated on all four types of graded materials. As for being caught cheating, only one person out of 205 (0.005%) who answered this question reported being caught cheating.

Hypothesis 2: First Time Cheating after COVID-19

The results supported Hypothesis 2. A large number of students cheated for the first time in online courses using online resources following the start of the COVID-19 pandemic in March 2020 across all types of graded materials. Overall, 46% (n = 98) of our sample reported cheating for the first time after COVID-19. Of those that reported cheating for the first time after COVID-19, 65.3% (n = 64) cheated for the first time on an exam, 60.2% (n = 59) cheated on a quiz, 37.8% (n = 37) cheated on homework, and 6.1% (n = 6) cheated on a project or paper.

We also calculated how many students cheated for the first time after the start of COVID-19 on one or more types of material. Of those that cheated for the first time after the pandemic, 51% (n = 50) cheated on one type of material, 31.6% (n = 31) cheated on two types of material, 15.3% (n = 15) cheated on three types of material, and 0.02% (n = 2) cheated on all four types of material.

Hypothesis 3: Use of Software to Prevent Online Cheating Using Online Resources

Hypothesis 3 was supported. As predicted, 85% of students (n = 182) reported that their instructors used software to prevent them from using online resources to cheat. In addition, 16.5% of students whose instructors used preventative software reported trying to circumvent this software.

Moreover, we asked participants what methods they have used to cheat despite the use of preventative software. Overall, out of 25 participants who answered this question, 48% (n = 12) reported taping their notes to the screen of their computer, 32% (n = 8) admitted using a separate device, 12% (n = 3) said they used a cheat sheet or wrote notes on their hands, and 8% (n = 2) reported using a friend to cheat (i.e., calling a friend during the test).

Hypothesis 4: Impact of Demographic Variables

Hypothesis 4 was not supported. There was no evidence that any demographic variables produced a main effect of decision to cheat, both overall (i.e., cheated or not) and across all graded material types (i.e., not cheated vs. cheated on one, two, three, or four graded materials).

Exploratory Analyses

We conducted four types of exploratory analyses. First, we examined the specific pattern of results described earlier based on Harbin and Humphrey (2013) view of cheating and opportunity (see Table 2). Using a repeated measures analysis of variance (ANOVA), we found a significant interaction between class format and type of material, F (3, 639) = 69.68, p < .001, η_p² = .25. For online classes, students reported that they had previously cheated (e.g., indicated yes to the question of if they had cheated) using online and offline resources on exams (M = .46, SD = .43), quizzes (M = .48, SD = .43), and homework (M = .49, SD = .44) at levels that did not differ reliably, (all t-tests < 1, ps > .36). In addition, cheating on these types of graded material were greater than that of projects/papers (M = .11, SD = .29; all t-tests > 10.70, ps < .001). For in-person classes, the pattern of results differed from that of online classes (see Table 2 for full t-test results). Students acknowledged cheating using online and offline resources on exams (M = .10, SD = .25), quizzes (M = .10, SD = .26), and projects/papers (M = .09, SD = .27) at levels that did not significantly differ (all t-tests < 1, ps > .54). Cheating on these types of graded material were lower than that of homework (M = .34, SD = .45; all t-tests > 8.30, ps < .001).

Table 2.

Comparisons Between Evaluation types for both Online Classes and in-person Classes

	Online Classes	In-Person Classes
Exams versus quizzes	t (213) = −0.90, p = .367, d =0.06	t (213) = .00, p = 1.000, d = .00
Exams versus homework	t (213) = −0.88, p = .380, d = .06	t (213) = −8.296, p < .001, d = .57
Exams versus projects/papers	t (213) = 10.72, p < .001, d = .73	t (213) = .548, p = .584, d = .04
Quizzes versus homework	t (213) = −0.23, p = .817, d = −.02	t (213) = −8.46, p < .001, d = .58
Quizzes versus projects/papers	t (213) = 11.48, p < .001, d = .78	t (213) = 0.61, p = .543, d = .04
Homework versus projects/papers	t (213) = 12.42, p < .001, d = .85	t (213) = 8.93, p < .001, d = .61

Note. Bonferroni adjusted alpha levels of .003 (.05/16) were used for each contrast.

Finally, the above patterns for each type of graded material on each type of class led to differences (see Figure 1) in the amount of cheating between online and in-person classes for exams, quizzes, and homework but there was no difference for projects/papers (t (213) = 1.71, p = .088). Specifically, online exams (M = .46, SD = .43) differed significantly from in-person exams (M = .10, SD = .25), (213) = 12.65, p < .001, d = .86; online quizzes (M = .48, SD = .43) differed from in-person quizzes (M = .10, SD = .25), t (213) = 12.97, p < .001, d = .89; and online homework (M = .49, SD = .44) differed from in-person homework (M = .34, SD = .45), (213) = 6.12, p < .001, d = .42. However, online projects/papers (M = .11, SD = .29) did not significantly differ from in-person projects/papers (M = .09, SD = .27), (213) = 1.71, p = .088, d = .12.

Figure 1.

Cheating as a function of class type and evaluation

Second, we examined frequencies of responses to Yes-No questions to better understand cheating behavior. This data indicated that: (a) very few students reported being caught cheating (.005% of students: one of 205), (b) 16.2% of students (34 of 203) stated that they would turn in someone who was cheating; (c) 64.8% (136 of 202) of students had been on a website or online group messaging platform (e.g., GroupMe) where they observed cheating but did not cheat (e.g., exited the site); and (d) 19% (40 of 200) of students stated that they told others not to cheat after they observed cheating.

The final exploratory analyses involved examining participants’ cognitive representations of how they felt after cheating or not cheating using networks of the participants’ mental models. Participants’ open-ended responses were imported into Pathfinder and frequency counts were made of each concept to identify the concepts most relevant in participants’ responses. After, a Pathfinder Network is created with the most relevant concepts and arranged in a way that highlights each terms’ importance. The most centrally located term is the most important term brought up in the responses. Each term, once in the network, is then referred to as a “node” which is connected to the other nodes by paths — the more paths connected to any single node, the more important that node is in the participants’ responses.

Figure 2(a) and 2(b) display the Pathfinder networks of participants’ open-ended responses to the question of why one cheated. For students who cheated the first time after the start of the COVID-19 pandemic the central node was “do not know the material” (see Figure 2(a)). This central node was attached to other reasons including “get good grades,” “no time,” “Covid,” “classes online,” “classes hard,” “not being prepared,” “stress,” and “hard to learn material.” Figure 2(b) shows the network for those who had cheated prior to the COVID-19 pandemic. Note that these students may have also cheated after the pandemic started. For these students, the central nodes were “stress” and “pressure to get good grades.” Other nodes included “no time,” “classes hard” and “not being prepared.” There is a noticeable difference between Figures 2(a) and 2(b), however, in that those who cheated for the first time after COVID-19 mentioned “Covid” while those who had cheated before did not.

Figure 2.

Pathfinder networks of why participants cheated (a) for the first time after the COVID-19 pandemic and (b) before the onset of the COVID-19 pandemic

Finally, Figures 3(a) and 3(b) present networks for how students felt after they had cheated. For students who cheated for the first time during the pandemic (Figure 3(a)) the central node was that they “felt good” about cheating. Other nodes included “felt guilty,” “felt bad,” “Covid,” and “felt fine,” Those who had cheated prior to the pandemic (Figure 3(b)) had a central node of “felt guilty” as well as nodes that were like those who cheated for the first time after COVID-19. For instance, Figure 3(b) included the nodes “felt bad,” “felt fine,” and “felt good.” Similar to Figure 2, Figures 3(a) and 3(b) differed in that those who cheated for the first time after COVID-19 mentioned “Covid” while those who had cheated before did not.

Figure 3.

Pathfinder networks of how participants felt after cheating (a) for the first time after the COVID-19 pandemic and (b) before the onset of the COVID-19 pandemic

Discussion

The current study presents novel and important findings. First, we found that most participants (approximately 75%) reported cheating across all types of graded materials (i.e., exams, quizzes, homework, and projects/papers). This finding is consistent with previous research investigating cheating behaviors (Bowers, 1964; McCabe et al., 2012). Second, the current COVID-19 pandemic led to an increase in first time cheating in online courses using online resources. Third, by investigating online cheating as a function of type of class (i.e., online vs. in person) and type of graded material, we have uncovered the nuances of cheating committed by college students. Put another way, college students are particularly savvy about when they cheat—it is not a simple case of cheating versus not cheating. Finally, we now have a clearer understanding of the reasons why college students decide to cheat, and how they feel about their decision to partake in this behavior.

The finding that cheating in college continues at such a high level is, of course, disconcerting, especially given that many colleges have Honor Codes (Yavorski, 2019), and instructors are using software developed to discourage cheating. However, the high percentage of cheating that we found may reflect changing in cheating behavior due to the current COVID-19 pandemic. Almost half of our participants acknowledged cheating after the pandemic began in March 2020. Moreover, these first time cheaters reported cheating most frequently on exams and quizzes. We expected an increase in cheating due to an increase in online classes and the increased impact of the pandemic on the mental health of college students (see Usher et al., 2021). Regarding the former, the move to a virtual learning environment offers greater opportunities for cheating (Harbin & Humphrey, 2013). Regarding the latter, mental health issues for young people likely led some students to view cheating as a necessary evil to cope with the demands of college. It is likely that the overall cheating numbers would have been lower than prior research (e.g., McCabe et al., 2012) if one assumed that a subset of those who for the first time after the pandemic would not have cheated if the pandemic had not occurred. Future research will be needed to determine if there was indeed a drop in cheating prior to the pandemic by continuing to investigate cheating behavior as the pandemic subsides to levels that allow more typical class schedules (i.e., many more in-person classes compared to online classes). However, the impact of the pandemic is likely to be long term such that the proportion of in-person and online classes may never be as disproportionate as it was prior to the pandemic (Friedman & Moody, 2021).

Concerning any decrease in cheating that might be observed in the future, we should note again that college instructors have been proactive in their efforts to prevent cheating. This was demonstrated in that most of the present participants reported that their instructors used software to prevent them from using online resources to cheat (supporting Hypothesis 3). However, our participants seemed one step ahead of their instructors. They reported bypassing this software by using their notes, a separate device, or relying on a friend to help them cheat. Whether efforts to prevent cheating in the future (including greater use of Honor Codes) will be successful awaits future study.

The specific pattern of cheating found across type of class (i.e., online vs. in-person) and type of graded material (i.e., exam, quiz, homework, and project/paper) offered a unique perspective on how students approach cheating. These factors have not been investigated together in any prior published study, although some studies have investigated cheating on different types of graded material (e.g., McCabe et al., 2012 included cheating on homework) and some have started to examine cheating in online classes (Harbin & Humphrey, 2013). Thus, the nuances of cheating in today’s virtual college classroom (especially post-pandemic) were unknown. The results showed clearly that students are more apt to cheat when the opportunity presents itself (Harbin & Humphrey, 2013), there are time constraints, and accuracy is important. Thus, cheating in online classes was higher for exams, quizzes, and homework compared to projects and papers, but for in-person classes cheating was only higher for homework compared to the other three types of graded material. Finally, cheating was higher for exams, quizzes, and homework in online classes compared to in-person classes. Future research will be necessary to better understand the decision-making process that students use to determine when cheating is a viable option.

In addition to investigating the patterns of cheating across critical variables, the present study offered a unique view regarding the mental representations of college students who cheat. Other studies have examined reasons for and feeling about cheating both in college (e.g., McCabe et al., 2012) and in non-college contexts (e.g., Peer et al., 2014), but the Pathfinder networks created using responses to open-ended questions offered a more detailed examination into students’ thoughts about deciding to cheat and their affectual response to cheating. On the one hand, we replicated prior research that has shown that students cheat because of feeling pressured and stressed about grades (Miller et al., 2017), and that cheating makes students feel guilty (DePalma et al., 1995). On the other hand, the networks offered new data about how students think and feel about cheating, including differences between students who cheated for the first time due to the COVID-19 pandemic and those who had already cheated in college. Students who cheated for the first time during the COVID-19 pandemic stated that they cheated because of the pandemic and taking online classes. It appears that students cheating for the first time found the material harder to learn because of the pandemic and their inability to prepare for their classwork. This also falls in line with previous research that online classes create more opportunities for students to cheat (Harbin & Humphrey, 2013). In addition, students who cheated for the first time during the pandemic stated that they “did not care” nor “did not feel bad” about cheating. We hope that future researchers investigating cheating will continue to have students generate responses to open-ended questions so that these responses can be used to construct Pathfinder analyses to investigate how students mentally represent cheating.

The present study did not find any effects of demographic variables or on- and off-campus activities on cheating. This included gender, working status, and fraternity or sorority affiliation. This finding was unexpected given the results of prior research (Danilyuk, 2019; McCabe et al., 2012). There are several possibilities for these non-significant effects including sample size and the sample not being representative of all college students (i.e., only included Psychology majors and minors). Of course, future research will be necessary to further investigate the impact of these variables and to determine whether changes in cheating have lessened the impact of demographics. For example, is it the case that cheating today is so rampant in college that it occurs regardless of demographic category?

Instructional Implications

As the online learning industry continues to grow (Miller, 2021), the opportunity and propensity to cheat will likely follow (Harbin & Humphrey, 2013). This is especially true during times when most in-person classes have switched to an online format, such as during the current COVID-19 pandemic. The current study has demonstrated such expectations, finding that almost half of the participants had cheated for the first time after the beginning of the pandemic. In fact, these findings were despite the efforts of instructors to prevent online cheating. Clearly, students are savvy enough to circumvent such efforts and are perhaps desperate enough for a good grade to do so.

College students are facing increasing pressure to do well in their classes. As an increase in jobs require (and will continue to require) at least a bachelor’s degree (Georgetown University, 2013), students face immense pressure to do well in class to earn a degree and eventually begin their careers. Further, students face pressure from their family, their peers, and society. Students may also place pressure on themselves because of wanting to avoid disappointing those around them. If students feel enough pressure to succeed, they may resort to cheating to do so. This conclusion was demonstrated in Figures 2(a) and 2(b), as students responded that they cheated due to stress and feeling that the material was too difficult to learn — both of which have likely been compounded by COVID-19. Distance learning likely increased the difficulty of learning already advanced material. In particular, students dealing technology difficulties such as internet connectivity issues may have been especially stressed during the COVID-19 pandemic, making the material harder to learn. Students’ professors may have also played a role in students’ stress and subsequent cheating. Professors’ abilities to translate in-person material to an online format may have played a particularly large role in the cheating behavior. This is supported by recent research (Usher et al., 2021), where almost half of participants reported they felt that instructional quality decreased due to the switch in class format.

The COVID-19 pandemic may also have brought about unique stressors, as students may have had to care for loved ones and worry about the health of both themselves and loved ones. Further, the exposure to the increasing detrimental effects of COVID-19 has been found to be related to decreases in mental health (e.g., increased anxiety and depression symptoms; Ni et al., 2020). Students were likely impacted by this, as news outlets and social media were continuously saturated with COVID-19 related news, such as the global and national death tolls, impacts on local communities, and instructions for best safety-practice during the pandemic.

Understanding the cheating behaviors of college students is important for several reasons. First, academic integrity is the cornerstone of academia. For that reason, understanding students’ propensity to cheat may help instructors and colleges alike prevent cheating. If successful, it may be the case that preventing cheating behavior is key to restoring academic integrity among students. To understand the best prevention methods, further research is necessary. Second, college often serves as professional training for students’ lives beyond graduation. Therefore, these years are a critical period for college students’ ethical development. If students interpret the ease of cheating in college as a way to excuse cheating or that cheating behavior is acceptable, then there could be serious implications for the rest of students’ lives, both personal and professional (i.e., in interpersonal relationships, careers, etc.). Lastly, as evidenced by the current findings, cheating behaviors are continuing to increase, in despite of instructors’ best prevention efforts. As higher education continues to rise in popularity, and technological advances are made to prevent cheating, students are savvy enough to adapt to circumvent these efforts. Having a better understanding of the underlying motivation to cheat is important for future prevention methods.

Conclusion

Limitations and Future Research

There are some limitations that should be acknowledged regarding the present study. First, as noted above the sample included Psychology majors and minors. It is possible that the nature of the Psychology curriculum and the elective courses taken by these majors impacted the cheating behavior of these students. We hope that future researchers will examine students in various majors to determine how the present results generalize across all college students. Second, undergraduate participants self-selected to be in the present survey. This can be viewed as a problem because it might be thought that a cheater would not want to admit to this “illegal” behavior. However, as was clear by the present results, having to acknowledge one’s cheating did not seem to prevent students (even those with relatively high GPAs) from participating. In fact, it could be argued that if a survey on cheating did lead to some students not participating, then the present results are an underestimate of the students who cheat. Third, related to the prior point, the response rate was low. However, this rate is consistent with the lower numbers reported by McCabe et al. (2012) and may reflect the concern we noted that some students (despite the anonymous nature of the survey) do not want to acknowledge any cheating behavior. Fourth, the current study utilized a survey that was created by the authors that did not include previously validated scales. While we wanted to capitalize on the small window of opportunity to study cheating behavior during COVID-19, our goal was not to develop a new cheating scale, but to get an understanding of cheating behavior during COVID-19. We hope future researchers continue to develop scales to measure cheating behavior. Lastly, it is important to mention that the survey was distributed at the end of the Spring semester, which may have impacted our results. The end of the semester can be a particularly stressful time for students, as final exams and plans for moving for the summer are stressful for students and may have had an impact on cheating behaviors.

To conclude, the results from the current study demonstrate that cheating behaviors in college is a pervasive issue. With the onset of the COVID-19 pandemic, cheating behaviors have likely worsened. Specifically, we found that three-fourths of our sample had cheated and that most of those who had cheated had done so on at least two types of graded material (e.g., exams and homework). We also found that COVID-19 had led many students to cheat for the first time in their collegiate career. This finding is most likely due to the increase of online classes and subsequent ease of cheating. Online classes provide students with a level of anonymity, allowing them to complete exams and quizzes without the direct supervision of an instructor. This, in combination with the additional stress of surviving a global health crisis, may present students with an unprecedented and extremely tempting choice to cheat. These results could be informative to educational institutions who are seeing an increase in online enrollment, especially as students continue to circumvent preventative software.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Open Practices

This article has received badge for Open Materials. More information about the Open Practices badges can be found at

ORCID iD

Mary M. Levi

References

Adzima

(2020). Examining online cheating in higher education using traditional classroom cheating as a guide. Electronic Journal of E-Learning, 18(6), 476–493. https://doi.org/10.34190/JEL.18.6.002

Bowers

W. J.

(1964). Student dishonesty and its control in college. Bureau of Applied Social Research, Columbia University.

Burgason

K. A.

Sefiha

Briggs

(2019). Cheating is in the eye of the beholder: An evolving understanding of academic misconduct. Innovative Higher Education, 44(3), 203–218. https://doi.org/10.1007/s10755-019-9457-3

Bustamente

(2020). Online education statistics. https://educationdata.org/online-education-statistics#summary

Carmichael

(2012). Harvard accuses 125 students of cheating. https://www.boston.com/news/local-news/2012/08/30/harvard-accuses-125-students-of-cheating

CBS News (2020). West point facing worst cheating scandal in decades. https://www.cbsnews.com/news/west-point-cheating-scandal-cadets-calculus-exam/

Danilyuk

(2019). Academic cheating statistics say there’s lots of work to do unicheck. https://unicheck.com/blog/academic-cheating-statistics

DePalma

M. T.

Madey

S. F.

Bornschein

(1995). Individual differences and cheating behavior: Guilt and cheating in competitive situations. Personality and Individual Differences, 18(6), 761–769. https://doi.org/10.1016/0191-8869(94)00208-A

Feder

(2020). Universities overcome bumps in transition to online teaching. Physics Today, 73(6), 22–25. https://doi.org/10.1063/PT.3.4492

10.

Friedman

Moody

(2021). Discover 14 current online learning trends. US News. https://www.usnews.com/higher-education/online-education/slideshows/discover-10-current-trends-in-online-education?slide=2

11.

Gamage

K. A. A.

Silva

E. K. D.

Gunawardhana

(2020). Online delivery and assessment during COVID-19: Safeguarding academic integrity. Education Sciences, 10(11), 301–325. https://doi.org/10.3390/educsci10110301

12.

Georgetown University (2013). Recovery: Job growth and education requirements through 2020. https://cew.georgetown.edu/cew-reports/recovery-job-growth-and-education-requirements-through-2020/

13.

Grijalva

T. C.

Nowell

Keryliet

(2006). Academic honesty and online courses. College Student Journal, 40(1), 180-185.

14.

Harbin

J. L.

Humphrey

(2013). Online cheating: The case of the emperor's clothing, elephant in the room, and the 800 lb. gorilla. Journal of Academic and Business Ethics, 7, 1–6. http://www.aabri.com/manuscripts/131450.pdf

15.

Harwell

(2020). Mass school closures in the wake of the coronavirus are driving a new wave of student surveillance. Washington Post. https://www.washingtonpost.com/technology/2020/04/01/online-proctoring-college-exams-coronavirus/

16.

Hylton

Levy

Dringus

L. P.

(2016). Utilizing webcam-based proctoring to deter misconduct in online exams. Computers & Education, 92–93, 53–63. https://doi.org/10.1016/j.compedu.2015.10.002

17.

Jenkins

B. D.

Golding

J. M.

Le Grand

A. M.

Levi

M. M.

Pals

A. M.

(2021). When opportunity knocks: College students' cheating amidst the COVID-19 pandemic. https://osf.io/8mkq4/?view_only=6357419258f34771b21ce49f4cd838df

18.

Kajackaite

Gneezy

(2017). Incentives and cheating. Games and Economic Behavior, 102, 433–444. https://doi.org/10.1016/j.geb.2017.01.015

19.

Kohlberg

(1971). 1. Stages of moral development as a basis for moral education. In Beck

C. M.

Crittendon

B. S.

Sullivan

E. V.

(Eds.), Moral education: Interdisciplinary approaches (pp. 23–92). University of Toronto Press.

20.

Lanier

(2006). Academic integrity and distance learning. Journal of Criminal Justice Education, 17(2), 244–261. https://doi.org/10.1080/10511250600866166

21.

McCabe

D. L.

(2005). Cheating among college and university students: A North American perspective. International Journal for Educational Integrity, 1(1). https://doi.org/10.21913/IJEI.v1i1.14

22.

McCabe

D. L.

(2020). Statistics. International Center for Academic Integrity. https://www.academicintegrity.org/statistics/

23.

McCabe

D. L.

Butterfield

K. D.

Trevino

L. K.

(2012). Cheating in college: Why students do it and what educators can do about it. The Johns Hopkins University Press.

24.

McCabe

D. L.

Trevino

L. K.

Butterfield

K. D.

(2001). Cheating in academic institutions: A decade of research. Ethics & Behavior, 11(3), 219–232. https://doi.org/10.1207/S15327019EB1103_2

25.

McGee

(2020). Texas A&M investigates 'large scale' cheating case in era of online classes. The Texas Tribune. https://abc13.com/texas-am-cheating-case-university-online-classes-scandal-aggie-honor-system-office-college-students-using-chegg/8823819/

26.

Miller

J. R.

(2020). Online cheating probes underway at Georgia Tech. Boston University. New York Post. https://nypost.com/2020/05/01/online-cheating-probes-underway-at-georgia-tech-boston-university/

27.

Miller

(2021). Online education statistics. https://educationdata.org/online-education-statistics

28.

Miller

A. D.

Murdock

T. B.

Grotewiel

M. M.

(2017). Addressing academic dishonesty among the highest achievers. Theory Into Practice, 56(2), 121–128. https://doi.org/10.1080/00405841.2017.1283574

29.

Murdock

T. B.

Anderman

E. M.

(2006). Motivational perspectives on student cheating: Toward an integrated model of academic dishonesty. Educational Psychologist, 41(3), 129–145. https://psycnet.apa.org/doi/10.1207/s15326985ep4103_1

30.

M. Y.

Yang

Leung

C. M. C.

Yao

X. I.

Wang

Leung

G. M.

Cowling

B. J.

Liao

(2020). Mental health, risk factors, and social media use during the COVID-19 epidemic and cordon sanitaire among the community and health professionals in Wuhan, China: Cross-sectional survey. JMIR Mental Health, 7(5), e19009. https://doi.org/10.2196/19009 Advanced online publication

31.

Peer

Acquisti

Shalvi

(2014). “I cheated, but only a little”: Partial confessions to unethical behavior. Journal of Personality and Social Psychology, 106(2), 202–217. https://doi.apa.org/doi/10.1037/a0035392

32.

Prihar

Wanna

(2019). Cheating at Yale: Students get honest about academic dishonesty. Yale News. https://yaledailynews.com/blog/2019/02/07/cheating-at-yale-students-get-honest-about-academic-dishonesty/

33.

Schvaneveldt

R. W.

(1990). Pathfinder associative networks: Studies in knowledge organization. In Ablex series in computational sciences. Ablex Publishing.

34.

Seipel

(2015). Stanford University looks into allegations of cheating by students. The Mercury News. https://www.mercurynews.com/2015/03/27/stanford-university-looks-into-allegations-of-cheating-by-students/

35.

Subin

(2021). How college students learned new ways to cheat during pandemic remote schooling. CNBC. https://www.cnbc.com/2021/03/21/how-college-students-learned-new-ways-to-cheat-during-covid-.html

36.

University of Kentucky (2020). Academic integrity. https://www.uky.edu/ombud/academic-integrity

37.

Usher

E. L.

Golding

J. M.

Han

Griffiths

C. S.

McGavran

M. B.

Brown

C. S.

Sheehan

E. A.

(2021). Psychology students’ motivation and learning in response to the shift to remote instruction during COVID-19. Scholarship of Teaching and Learning in Psychology. Advance online publication. https://psycnet.apa.org/doi/10.1037/stl0000256

38.

Varvel

V. E.

(2005). Honesty in online education (Vol. 6). https://www.uis.edu/ion/wp-content/uploads/sites/224/2018/09/VarvelCheatPoint2005.pdf

39.

Watson

Sottile

(2010). Cheating in the digital age: Do students cheat more in online courses? Online Journal of Distance Learning Administration, 13(1), 1–9.

40.

Whitley

B. E.

Nelson

A. B.

Jones

C. J.

(1999). Gender differences in cheating attitudes and classroom cheating behavior: A meta-analysis. Sex Roles, 41, 657–680. https://doi.org/10.1023/A:1018863909149

41.

Yavorski

(2019). The increasing relevance of honor codes. Higher Ed Connects. https://higheredconnects.com/honor-codes/

42.

Young

J. R.

(2012). Online classes see cheating go high-tech. The Chronicle of Higher Education. https://www.chronicle.com/article/online-classes-see-cheating-go-high-tech/

43.

Glanzer

P. L.

Sriram

Johnson

B. R.

Moore

(2017). What contributes to college students' cheating? A study of individual factors. Ethics & Behavior, 27(5), 401–422. https://psycnet.apa.org/doi/10.1080/10508422.2016.1169535