Systematic review and meta-analysis as a structured platform for teaching principles of experimentation

摘要

Access to knowledge has never been easier in the internet age, and so it is important that students develop skills to discriminate undependable information from reliably investigated research. We have created an exercise that teaches good research practice by exploring the history, ethics, and design of clinical trials. Students apply their understanding of these principles through an assessed systematic review and meta-analysis (SRMA) exercise. Here, a clinically themed hypothesis is tested using a structured literature search in conjunction with an eligibility matrix to map study design, ethics, subject selection, randomization and blinding, methodological standards, study power, and other potential sources of interstudy heterogeneity. Data extracted from selected studies are used to produce a forest plot with an aggregated effect size, confidence range, and measure of interstudy heterogeneity. A funnel plot is then used in conjunction with the eligibility matrix to evaluate study bias tendency, and, in this way, students reflect on the factors that promote disparate conclusion-making among studies with a common research focus. This exercise produced a normally distributed grade-profile across three academic-year cohorts, and comparison of individual exercise grade with year-long aggregated average suggested students who performed less well on conventional assignments engaged successfully with the systematic nature of this assessment. Those opting to use this format for their final-year capstone project also performed above their grade point average from the preceding year. We suggest that SRMA offers a readily applied method for students to quantitatively explore how differences in experimental research practices influence study dependability.