Confounding and Interaction: An Interactive Self-Study Quiz

Abstract

Abstract Introduction The study of evidence-based medicine involves complex statistical concepts. One of the most challenging aspects for students is the transition from simplified explanatory examples of such topics to analysis of their application in clinical research. This module bridges that gap and guides the progression from following step-by-step examples to independently approaching open-ended questions. It is designed for use by medical and other professional students to reinforce the specific topics of confounding and interaction. Methods Using examples taken from medical literature, the problem sets emphasize recognition of potential confounders, stratification to identify instances of confounding, and study design that minimizes risk of confounding. At every step, learners have the option to follow a link for applicable reference information. Therefore, students attempt to solve a problem independently, but if at any point they are stumped, they can immediately access progressively more detailed reminders of the core concept required to solve the problem until they reach the point from which they are able to approach the rest of the problem independently. This module is most effective as a self-study tool to assess and promote understanding. Results The module was provided to Emory University School of Medicine first-year students to supplement their classroom-based course in evidence-based medicine. The students used the module as a self-study tool for the final exam as well as for preparation for future practical application. The final average examination grade for this course improved 5 percentage points from the previous average. Discussion We also distributed this module to fourth-year medical students enrolled in an evidence-based medicine elective in spring 2014. A limitation of the module is that it is not designed as a primary resource for teaching the included material. Expanding the resource in the future to include more basic information may facilitate its use as an independent-learning tool.

Keywords

• Clinical Research
• Evidence-Based Medicine
• Biomedical Research
• Bias
• Study Design
• Stratification