High-Stakes Collaborative Testing: Why Not?

Phenomenon: Studies of high-stakes collaborative testing remain sparse, especially in medical education. We explored high-stakes collaborative testing in medical education, looking specifically at the experiences of students in established and newly formed teams. APPROACH: Third-year psychiatry students at 5 medical schools across 6 sites participated, with 4 participating as established team sites and 2 as comparison team sites. For the collaborative test, we used the National Board of Medical Examiners Psychiatry subject test, administering it via a 2-stage process. Students at all sites were randomly selected to participate in a focus group, with 8-10 students per site (N = 49). We also examined quantitative data for additional triangulation. FINDINGS: Students described a range of heightened emotions around the collaborative test yet perceived it as valuable regardless if they were in established or newly formed teams. Students described learning about the subject matter, themselves, others, and interpersonal dynamics during collaborative testing. Triangulation of these results via quantitative data supported these themes. Insights: Despite student concerns, high-stakes collaborative tests may be both valuable and feasible. The data suggest that high-stakes tests (tests of learning or summative evaluation) could also become tests for learning or formative evaluation. The paucity of research into this methodology in medical education suggests more research is needed.

Team cohesiveness, team size and team performance in team-based learning teams.

OBJECTIVES: The purpose of this study was to explore the relationships among variables associated with teams in team-based learning (TBL) settings and team outcomes. METHODS: We administered the National Board of Medical Examiners (NBME) Psychiatry Subject Test first to individuals and then to teams of Year three students at four medical schools that used TBL in their psychiatry core clerkships. Team cohesion was analysed using the Team Performance Scale (TPS). Bivariate correlation and linear regression analysis were used to analyse the relationships among team-level variables (mean individual TPS scores for each team, mean individual NBME scores of teams, team size, rotation and gender make-up) and team NBME test scores. A hierarchical linear model was used to test the effects of individual TPS and individual NBME test scores within each team, as well as the effects of the team-level variables of team size, team rotation and gender on team NBME test scores. Individual NBME test and TPS scores were nested within teams and treated as subsampling units. RESULTS: Individual NBME test scores and individual TPS scores were positively and statistically significantly (p < 0.01) associated with team NBME test scores, when team rotation, team size and gender make-up were controlled for. Higher team NBME test scores were associated with teams rotating later in the year and larger teams (p < 0.01). Gender make-up was not significantly associated. CONCLUSIONS: The results of an NBME Psychiatry Subject Test administered to TBL teams at four medical schools suggest that larger teams on later rotations score higher on a team NBME test. Individual NBME test scores and team cohesion were positively and significantly associated with team NBME test scores. These results suggest the need for additional studies focusing on team outcomes, team cohesion, team size, rotation and other factors as they relate to the effective and efficient performance of TBL teams in health science education.

Team Emotional Intelligence, Team Interactions, and Gender in Medical Students During a Psychiatry Clerkship.

OBJECTIVE: This study examined the relationship between team emotional intelligence, quality of team interactions, and gender. METHODS: Psychiatry clerkship students participating in Team-Based Learning (TBL, n = 484) or no TBL (control, n = 265) completed the Workgroup Emotional Intelligence Profile (WEIP-S) and the Team Performance Scale (TPS). RESULTS: Significant correlations (p < 0.01) existed between quality of team interactions (i.e., TPS) and team emotional intelligence (i.e., WEIP-S) subscales, but not gender. Control and TBL groups experienced significant increases in WEIP-S subscales pre to post (p < 0.01, η (2) = .08), with the TBL group experiencing significantly higher gains in three of four subscales. Control group scored higher on TPS. CONCLUSIONS: A significant relationship exists between team emotional intelligence and quality of team interactions. Gender was unrelated to TPS or WEIP-S subscales. TBL group experienced higher gains in WEIP-S subscales while the control group experienced slightly higher TPS scores. Results suggest implications for medical educators who use TBL.