Investigating the existence of social networks in cheating behaviors in medical students

Abstract

BACKGROUND: Most studies on academic cheating rely on self-reported questionnaires and focus on the individual, overlooking cheating as a group activity. The aim of this study is to estimate the true prevalence of cheating/anomalies among medical students using a statistical index developed for this purpose, and to explore the existence of social networks between anomalies in students' results. METHODS: Angoff's A index was applied to a sample of 30 written examinations, with a total of 1487 students and 7403 examinations taken, from the 2014/2015 academic year of the Faculty of Medicine of the University of Porto to detect anomaly pairs. All analyses are within the same academic year and not across years. Through simulations, the sensitivity and specificity of the statistical method was determined, and the true prevalence of anomalies/cheating was estimated. Networks of anomaly pairs were created to search for patterns and to calculate their density. RESULTS: The percentage of students who cheated at least once increased with the year of medical school, being lowest in the first year (3.4%) and highest in the fifth (17.3%). The year of medical school was associated with anomalies (p < 0.05). The network's density was also lowest in the first year (1.12E-04) and highest in the fifth (8.20E-04). The true prevalence of anomalies was estimated to be 1.85% (95%CI: 1.07-3.20%). CONCLUSIONS: These findings suggest that some students are involved in social networks of cheating, which grow over time, resulting in an increase of anomalies/cheating in later academic years.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 20 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.