Two-Stage Exams


This is a reflection on a two-stage exam video published to YouTube on November 10, 2014 by the University of British Columbia (“UBC”) channel “ubcpublicaffairs”.


Upon viewing the video, I immediately thought it was an incredibly innovative and interesting approach to getting students involved in their own learning.  I especially enjoyed how the students draw on each other’s knowledge and teach each other during the second stage of the two-stage exam.  I do wonder how this might effect a poor student, and whether or not this process would be deemed fair by a good student.  My interest was definitely piqued in a big way.  


I had never heard of two-stage exams prior to watching this video.  The key insight that I now have is the possibility of increasing subject retention as a result of the two-stage exam — an issue that I struggle with in my teaching.  Another key point was reduced exam stress for students, which I know would be incredibly helpful for some of my classes.  Because these key insights are of great importance to me in my teaching, I should investigate them further and figure out how I might be able to incorporate the two-stage exam technique in my teaching.

Regarding one of my main concerns – how the two-stage exam would be perceived by my students, a paper titled, “Two-Stage Exams” published in October 2014 by the Carl Wieman Science Education Initiative, addresses this matter.  According to this paper, “during the group part [of the exam] students receive immediate, targeted feedback on their solutions from their fellow students and see alternative approaches to the problems.  This makes the exam itself a valuable learning experience while also sending a consistent message to the students as to the value of collaborative learning” (para. 1).  The paper also states that student response to the use of two-stage exams was overwhelmingly positive: 87% of students surveyed recommended continued use of two-stage midterm exams while only a few percent recommended against their use.  The fact that the exam process was first introduced in the UBC Faculty of Science in 2009, and is now being used in at least 20 science courses,  addresses any hesitation I had about introducing this technique to my students.

Concerning student retention, Gilley and Clarkston (2014) note the previous work of Cortright et al. (2003), who “reported significantly greater retention after 4 weeks on questions students had answered in groups compared with questions answered as individuals” (p. 84).  While those results have been contested, Gilley and Clarkston (2014) showed a statistically significant improvement in student performance on an individual retest after a group test had been conducted.  Knierim, Turner, and Davis (2015) found additional benefits in the classroom as a result of implementing two-stage exams, such as significantly increased attendance rates throughout the semester, which may have been another reason, in addition to the collaborative testing, that overall grades improved.  They also note that lower-achieving students benefited the most from the collaborative component of the exam process.


There is now little doubt that implementing collaborative testing through two-stage exams in my courses would be a positive step.  The most sensible approach would seem to be creating a mid-term exam in this format.  I plan to implement a two-stage exam in at least one course this upcoming semester.  I turn back to the paper titled, “Two-Stage Exams” by the Carl Wieman Science Education Initiative to help address how I can implement this style of collaborative testing, as the paper outlines several strategies that have worked well.  

The first recommendation is to explain to the students why the exam will be conducted using this format, and to do so on the first day of classes.  This is also a good opportunity to explain fairness, and that a student’s grade will not go down due to the group portion of the exam.  A policy that the student’s grade cannot be lower than the individual score, will help address fairness.  

Regarding exam logistics and administration, it is recommended that the exam be scheduled for only ⅔ the time that would be given for a normal exam.  Then, ⅔ of the scheduled time would be provided to complete the individual portion of the exam, with the remaining portion provided for the collaborative portion.  The majority of the exam grade should be for the individual portion and I like the idea of 85%, with the remaining 15% of the exam grade for the collaborative portion.    

During the exam, it is critical to give clear instructions during the transition between the individual and group portions, though my smaller class sizes will relieve some of the difficulties of the transition.  It is suggested that the group size be 3 or 4 students and in my first attempt I will assign the groups by the order of their student numbers.  While there are various approaches to the group portion of the exam, such as repeating a subset of questions, or adding more challenging questions, my first step will be to simply repeat the same exam.

Overall, it does not seem incredibly complicated to implement a two-stage exam in one of my courses and I hope that me and my students experience even some of the noted benefits of conducting collaborative testing.


Carl Wieman Science Education Initiative, (2014).  Two-stage exams.  Retrieved from

Gilley, B.H., Clarkston, B., (2014).  Collaborative testing: Evidence of learning in a controlled in-class study of undergraduate students.  Journal of College Science Teaching, Vol. 43, No. 3, 83-91.

Knierim, K., Turner, H., Davis, R.K., (2015).  Two-stage exams improve student learning in an introductory geology course: Logistics, attendance, and grades.  Journal of Geoscience Education, 63, 157-164.