Two-stage exams literature round-up

Two-stage exams are exams which students first write individually before getting into groups to complete the same (or similar) set of questions again. This harnesses students’ engagement in the high-stakes environment to create a learning opportunity through peer interactions and immediate feedback. Besides, it’s fun.

Over the first couple months of the summer semester, @jossives and I have been leading a reading group we affectionately refer to as GIRG—the Group Interactions Reading Group. (My first naming attempt was GEGRG—Group Exams Group Reading Group—but there were too many “Group”s involved.)

This is a quick round-up post of the literature we’ve looked at so far, and some of the main points or questions that came out of these for us.

Group quizzes as a learning experience in an introductory lab

Brett Gilley and Sara Harris. (Poster available here.)

Summary: Performance data from solo and group phases show that groups do well (and self-correct) except in the case where everyone has the same individual wrong answer.

Points of interest/extension:

  • “Discriminating questions foster discussion.” This might be useful for test design.
  • What behaviours do students engage in during the group stage?

Does group composition impact group scores in two-stage collaborative exams?

Sonya Sabourin, Tamara Kelly, and Colin Montpetit. (Poster available here.)

Summary: Aims to build a predictive model of group performance based on the different factors that characterize groups.

Points of interest/extension:

  • Big question: What’s the most useful way to characterize groups? May need some measure of group ability (individual average or individual score of best student) and the heterogeneity of the group’s abilities.
  • Does having students choose their own groups mitigate some of the common problems with assigned groups? (Such as having students feel isolated.)

Collaboration or copying? Student behavior during two-phase exams with individual and team phases

Ian Beatty.

Summary: Creative perspectives on performance data from solo and group phases indicate that students do engage in meaningful and productive debate during the group phase.

Points of interest/extension:

  • “From-nowhere” group responses—when the group selects an answer that no individual had—would be a particularly interesting subset of the data to explore, perhaps with case studies.
  • The “team bonus” Beatty gives—a bonus in grades for gains your groupmates make from the solo to the group phase—might encourage team behaviours.
  • The data examined in this paper are from difficult tests: “Even the best students in the course rarely earn more than 75% of the credit during the solo phase.” How does test difficulty impact learning?

Impact of peer interaction on conceptual test performance

Chandrealekha Singh. (Available also on arXiv.)

Summary: Some of the students wrote a two-phase test as solo->pair, some as pair->solo, and a control group just wrote solo. Comparison on a retention test showed similar performance for both ‘pair’ conditions, and worse performance for the control group.

Points of interest/extension:

  • Similar retention for both solo->pair and pair->solo; perhaps committing to an answer is less important than peer interactions?
  • In ~30% of cases where both students of a pair were incorrect on the solo phase, they got the question correct on the group phase. (Co-construction of knowledge.)
  • Similar gains seen for students no matter the group type (whether paired with a student of similar or different ability).

Effective student teams for collaborative learning in an introductory university physics course

Jason Harlow, David Harrison, and Andrew Meyertholen.

Summary: Examined the possible effect of group metrics (in tutorial groups) on student learning, as measured by FCI and exam performance. Found no effect of group composition (spread of abilities versus all similar abilities), size of group (three versus four), and for groups with an isolated female student.

Points of interest/extension:

  • No evidence for the common wisdom that groups should include students with a spread of abilities.



One comment

  1. Yay. Jared did our homework for us.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: