Efficacy of Collaborative Testing for Long‐Term Retention of Medical Knowledge

Abstract

Collaborative testing is a trending tool in higher education where 2–3 students are grouped together to complete an exam (Rao, Collins & DiCarlo, 2002). Research has shown that collaborative testing enhances student learning and retention when compared to individual testing (Cortight, Collins, Rodenbaugh, & DiCarlo, 2003). However, the existing literature does not address how collaborative testing affects long‐term knowledge retention, which is critically important for medical students who must retain knowledge for clinical application and board exams. Mixed methods were used to study the effects of collaborative testing on long‐term knowledge retention, as well as how students approach and perceive collaborative testing.MethodsFirst‐year medical students were tested on basic science knowledge through two renal course exams spaced 1.5 weeks apart. For each exam, questions were equally divided into two conditions: “collaborative testing” and an “exam review” group. Both testing conditions consisted of groups of 2–3 students. For collaborative testing, groups discussed and answered exam questions after completing the exam individually. However, the correct answers were never revealed. The exam review groups reviewed exam questions with the correct answers provided. To rule out a temporal effect, the students were randomly divided into two groups. The exam was also divided in half. Group A took their collaborative questions immediately after the individual exam followed by the exam review of questions with answers, and Group B experienced the two conditions in the opposite sequence. To assess long‐term knowledge retention, students took a post‐test of basic science renal content at the start of their clinical renal course 8 months later. 20 questions with strong point biserial values were selected, 5 each from the collaborative and exam review conditions of the two renal exams. Questions selected from the two conditions covered similar content and had similar student performance (Table 1). Knowledge retention was computed, and a paired t‐test was used to compare retention values between the two conditions. Qualitative data included student responses to a survey on process and perceptions of the conditions and was collected at the mid‐point of the first year renal course. A grounded theory approach was used to identify themes in student responses.Results and DiscussionWhile there was no significant difference in performance between the two conditions on the initial exams, the post‐test administered 8 months later showed a significantly higher score on the collaborative testing questions as compared to exam review questions. This data suggests an improved long‐term retention of medical knowledge for collaborative testing without answers provided (Figure 1, Table 2). Thematic analysis of qualitative data from surveys indicated that students engaged in much more discussion in the collaborative condition. However, they expressed the desire of having answers provided and that certainty of correct answers was critical to their learning. Based on survey data, future research would include merging the exam testing strategies by having group testing work through questions and reveal the correct answer with a scratch‐off sheet. This new approach could provide insight on how discussion and decision‐making plays a role in long‐term retention. Finally, this study took place in an integrated curriculum and has relevance for anatomists in many teaching environments.This abstract is from the Experimental Biology 2018 Meeting. There is no full text article associated with this abstract published in The FASEB Journal. The 20 questions from the initial exams that were selected to create the post‐test exam 8 months later are statistically similar quality via their point biserial (.19 –.45; mean=.33) and content. The questions also yielded similar initial performance (mean of 77.94% for collaborative and 76.18% for exam review). Each student answered the same question twice ‐‐ once during the initial exam and a second time during the post‐test exam 8 months later. Group Content Point biserial Initial performance Post‐test performance Collaborative A MICRO 0.46 0.8 Initial mean=77.94% 0.46 Post‐test mean = 49.80% PHARM 0.39 0.75 0.64 ANAT 0.39 0.96 0.46 PHYS 0.38 0.71 0.49 HISTO 0.36 0.88 0.64 B PHYS 0.44 0.8 0.43 PHYS 0.3 0.75 0.71 PHYS 0.31 0.96 0.53 PHARM 0.33 0.71 0.19 BIOCHEM 0.3 0.88 0.43 Exam Review A HISTO 0.45 0.84 Initial mean = 76.18% 0.41 Post‐test mean = 43.90% PHYS 0.39 0.68 0.42 ANAT 0.19 0.94 0.7 MICRO 0.2 0.69 0.44 PHARM 0.26 0.74 0.55 B PHYS 0.26 0.73 0.33 PHYS 0.38 0.53 0.09 PHYS 0.32 0.72 0.69 BIOCHEM 0.24 0.89 0.12 PHARM 0.24 0.76 0.64 Mean = .33 The significantly higher score for collaborative testing after 8 months suggests that this condition yields higher long‐term retention for medical knowledge. N Mean Mean difference Standard deviation Paired t‐test P value Significance Initial Collaborative 132 .78 6.11 30.29 2.32 0.02 <0.5 Exam Review .76 Post‐test Collaborative 132 .50 1.76 16.89 1.19 .23 <0.5 Exam Review .44