What kind of test can engage students and enhance their learning? Participants in a Minnesota Campus Compact workshop last Friday experienced firsthand a team-based technique that was fun and informative. Individuals first completed a multiple-choice test on their own, then gathered in small groups to discuss the questions and determine the group’s answers together. By next going through and scoring both our individual and group tests, we got immediate feedback, learned the right answers—and found that no one individual had scored higher than the group. We’d also enjoyed the opportunity to share the reasoning behind our answers and learn from others’ perspectives. The next step was to apply what we’d learned, which in this case meant applying principles about designing courses for significant learning to revising specific course syllabi.
Faculty around the country have conducted research on this kind of team testing, with similarly positive results. Dr. Robert Dunbar at the University of Minnesota Rochester has begun related research and shares his preliminary findings and both pedagogical and research motivations below:
The group component [in classes in this study] goes beyond testing. Students complete study guides prior to class. Then in class, they complete a quick “confusion and clarifications” survey (think – muddiest point) before they do anything else. Next, they split into their groups to discuss the study guide questions/concepts and then complete the same “confusions and clarifications” survey. The second time, they focus on highlighting concepts that were not clarified by the small group discussion. We then take the results of the second survey to guide the focus of our “lecture”/class discussion. Therefore, the students help to clarify topics in their groups before we do anything. Preliminary results of the surveys suggest that there is a significant gain in understanding just through peer interactions before we lecture. This is not yet published and is still very preliminary.
Pedagogy – Based on feedback from graduate/professional programs as well as industry, it is clear that there is a demand for graduates who can effectively work in groups. However, we also still needed to support and encourage individual, self-accountability. How do we reconcile these apparently opposing goals? Furthermore, can we encourage people to value group work AND self-accountability without generating animosity between high and low performers? The model that I (and the student based faculty who work with me) have implemented includes a high point value for the individual exams (~150 pts) and a lower possible extra credit contribution from the group tests (up to 10 pts). Under this model, performance of the group does seem to be influenced by the high performers but the high performers also value the contributions of all members of the group to make up the gaps in understanding that they (the high performers) have. Furthermore, the lower performers no longer resent the high performers as “curve breakers” because there is no curve. Rather, the extra-credit earned by all effectively replaces any curve so ALL students benefit when the group performs better.
Research – Does working in groups facilitate learning the material? This is a tricky question. The data collected to date argues that the vast majority of the time, the group result is above even the highest performer in the group. In other words, even the highest performers benefit from working in groups. As with all studies of this nature, there are exceptions but these are rare. Furthermore, and this is anecdotal for the moment, students actually discuss and appear to learn while they go over the test as a group. I am currently trying to figure out how gender and social self-efficacy relate to group performance as well as how to capture the level of learning that appears to be happening. Undoubtedly, future analysis will include questions that appear on multiple exams and, someday with appropriate IRB approval, an analysis of student discussions while they work.
While this team-oriented approach to traditional course content seems like a natural fit for classes that form teams to complete community-engaged projects, it is applicable to a wide array of courses. Harvard physicist Eric Mazur developed a similar peer instruction technique that has received widespread acclaim and adoption in the sciences. His research and UMR’s unique curriculum are both highlighted in an American Radio Works documentary aired last week, Don’t Lecture Me. A few additional resources on this topic:
Even beyond the context of teaching and course development, the research showing that student teams do better than individuals is intriguing. In The Wisdom of Crowds: Why the Many Are Smarter Than the Few and How Collective Wisdom Shapes Business, Economies, Societies and Nations (Doubleday, 2004), James Surowiecki argues that groups can make better decisions or predictions than experts, particularly when those groups draw on a diversity of opinions from individuals whose opinions were reached independently, and when decentralization allows the group to draw on local knowledge to determine their collective position. Given how important collaborative action and decision-making are in a democracy, it’s exciting to speculate that positive experiences with team-based learning could also increase people’s inclination to engage with others to address important public issues. Anyone interested in a research project?