Group Testing in Medical Education: An Assessment of Group Dynamics, Student Acceptance, and Effect on Student Performance

Abstract

Our institution uses a model of group testing in which students retake exam questions in a group after submitting individual answers. We examined the effects of group testing on exam performance and group dynamics. Students scored higher after discussing the questions. When the same cohort of students was allowed to submit an independent answer one year and then required to achieve consensus the next, the improvement in their performance after discussion did not change. 83.1% of students reported teaching during the group exam. 65.1% reported groups achieve unanimity "often"; 56.4% reported feeling pressured by peers. While in preformed groups, personal stress was decreased in 44% of respondents and there was less conflict in 60% of respondents compared to randomly formed groups. Hostility was three times more likely in a random group than a pre-formed group (37% vs. 12%). Both teaching and seeking clarification occurred more often in pre-formed groups. These data suggest that group testing can be used to enhance student performance, understanding, and retention. However some students feel pressured and find these groups to be hostile environments. Testing within a pre-formed group may provide students with an opportunity to teach each other with decreased stress and anxiety.

Exploratory Factor Analysis of a survey on group-exam experiences and subsequent investigation of the role of group familiarity

“…In addition to targeting improvements in intra-group familiarity, we also sought to improve equity in participation and decision-making in this intervention through the sharing sharing of curated responses to a previous survey item: "Provide advice you would give to a future student in this class to get the most out of their group exam experience". Although we have not yet analyzed the results of this pilot intervention, an example from group testing in medical education [4] supports the plausible efficacy of the familiarity component of this intervention.…”

“…The literature on student experiences with in-class groupwork informed our survey development related to themes of group familiarity [3,4], participation equity [3], productive group-work behaviours [4,5], comfort [3], and gender [6] and other social identities [7].…”

Improved Learning Through Collaborative, Scenario-based Quizzes in an Undergraduate Control Theory Course

“…Past research on collaborative quizzes has generally sought experimental control and used random group assignment. However, in one study that contrasted random versus selfselected assignment, students in self-selected groups reported lower rates of conflict and stress, compared to their randomly assigned peers [95]. In a similar study, students who self-selected tended to report greater agreement that this approach improved their individual learning, whereas those randomly assigned tended to report greater agreement that the approach improved their collaboration skills [96].…”