Monitoring methods and student performance in distance education exams

Abstract

PurposeThe objective of this study was to compare three distinct methods of test monitoring in the context of distance education—non-proctored remote or online tests, traditional face-to-face proctored tests and remotely proctored tests using software—to analyze if the method in which tests are monitored influences the obtained grades.Design/methodology/approachThe experiment was carried out at the postgraduate level in the Master's Degree in Administration program in the modality of distance education, with a total of 296 students during three terms wherein the monitoring method of the final exam varied, keeping the other variables constant. This study used a quantitative method in which the distribution of grades was analyzed; and the grades from each method were tested. Finally, using a multiple linear regression model with dichotomous variables, the impact on students' academic performance with each method was quantified.FindingsThe results indicated that the remotely proctored online test grades were seven points lower with respect to the traditional method. This result does not mean that the lower scores in the remote proctored condition were due to better adherence to academic honesty, maybe this could be due to test anxiety, technology interference or a number of other factors that would confound the validity of the final test score.Practical implicationsThe results indicated that the non-proctored online test favored the grade in four points with respect to the traditional method.Social implicationsThe authors conclude to support recommending non-proctored online test, this can be a closer substitute to the traditional method than remote application with software monitoring.Originality/valueNot exist another paper to compare three distinct methods of test monitoring in the context of distance education.

Student Cheating Detection in Higher Education by Implementing Machine Learning and LSTM Techniques

“…In order to identify cheating during an online test, the authors of [5] developed a system for assessing the head position and time delay. A high statistical correlation between cheating activity and a student's head position change relative to a computer screen was also discussed.…”

“…We believe that additional staff is sufficient to oversee students remotely rather than automatically in order to supervise more students [3]. Although this is theoretically conceivable, it cannot be scaled for a course when more students are sitting their face-to-face or distance test [4][5][6]. Complex logistics would be necessary for this.…”

Instructor Perceptions of e-Proctoring Software

“…Academic integrity is an essential factor for education institutions due to its impact on institutional reputation [60]. However, studies examining students' performance and academic dishonesty in online remote proctored formats compared to in-person formats have found mixed results [22,24,33,38,42,51,66], suggesting a lack of consensus on the efficacy of online remote proctored exams. Motivations and mechanisms for dishonest academic behaviour have been widely studied [60].…”

“…There are key distinctions between the act of monitoring students during in-person proctoring compared to eproctoring. While in-person proctoring typically administers testing in a controlled environment such as an examination hall with human proctors, e-proctoring often involves test completion in uncontrolled environments, with a wide range of technologically-mediated monitoring capabilities [66,72]. Technological features of modern e-proctoring systems offer an enhanced capability to detect/deter cheating using not just well-known measures such as video and audio monitoring, but also using novel methods of monitoring to account for the potential for academic dishonesty introduced by the online remote format [60,66,72].…”

“…Rodríguez-Villalobos et al, [66] conducted an experimental study comparing three different test monitoring methods in the distance education context (i.e., non-proctored online remote tests, traditional face-to-face proctored tests, and online remote proctored tests using software monitoring) to determine if the test monitoring method influences test grades. They found all three groups to be significantly different: online remote proctored test grades were significantly lower than traditional face-to-face proctored tests, not necessarily because remote proctoring limited cheating.…”