Proposal Title

Observed Cheating and the Effects of Random Seat Assignment

Track

Research Proposal / Assessment of Student Learning

Proposal Abstract

Using a multiple choice exam design that allows us to empirically measure the degree to which students copy answers from their peers, we conduct an experiment to determine whether random seat assignment deters cheating relative to a control group of students allowed to choose their seats. Our results demonstrate a significant decline in measured cheating within the assigned seating sample. This study contributes to the literature by providing a measurement of actual cheating frequency among students, as opposed to relying on reported cheating in anonymous surveys, and by demonstrating that an easily implemented deterrent can significantly reduce instances of cheating.

Proposal Description

Using the multiple choice exam design described in Fendler and Godbey (2015), we record and analyze the cheating behavior of undergraduate students on a corporate finance final exam at a large state university. We conduct a controlled experiment where students in one classroom were unexpectedly given a random seating chart prior to the exam, while students in a second classroom were allowed to seat themselves. The two classrooms were nearly identical in student composition, classroom layout and proctoring during the exam.

While our results demonstrate cheating behavior in both classrooms, we find that students in the classroom without assigned seats had a significantly higher frequency of responses copied from neighboring papers. These results are statistically significant at the 99% confidence level and are present when the analysis focuses on classroom-wide or student-specific cheating.

Dating back at least to Parr (1936), cheating among higher education students has been a popular topic in the fields of education and psychology. Because student cheating is generally unobservable, most empirical work in this area relies on anonymous student survey data; however, it is difficult to draw concrete conclusions about the true prevalence of cheating from these surveys, as there can be large discrepancies between reported and actual behavior (Nelson and Schafer 1986). Most of the literature that has attempted to document actual cheating has followed the research design of Parr, which highlights instances of individual opportunistic cheating. Our test design is novel in that it allows us to identify instances of active social cheating (i.e., copying a neighbor’s answer), a very common form of cheating according to survey data.

In addition to allowing us to quantify actual instances of copying, our experimental results have prescriptive implications for educators, demonstrating that a simple and relatively costless deterrent can significantly reduce cheating behavior.

Session Format

Presentation Session

Location

Room 5

Publication Type and Release Option

Presentation (Open Access)

Share

COinS
 
Mar 30th, 4:00 PM Mar 30th, 4:45 PM

Observed Cheating and the Effects of Random Seat Assignment

Room 5

Using a multiple choice exam design that allows us to empirically measure the degree to which students copy answers from their peers, we conduct an experiment to determine whether random seat assignment deters cheating relative to a control group of students allowed to choose their seats. Our results demonstrate a significant decline in measured cheating within the assigned seating sample. This study contributes to the literature by providing a measurement of actual cheating frequency among students, as opposed to relying on reported cheating in anonymous surveys, and by demonstrating that an easily implemented deterrent can significantly reduce instances of cheating.