More Evidence on the Use of Constructed-Response Questions in Principles of Economics Classes

This study provides evidence that constructed response (CR) questions contribute information about student knowledge and understanding that is not contained in multiple choice questions (MC). We use an extensive data set of individual assessment results from Introductory Macro- and Microeconomics classes at a large, public university. We find that (i) CR scores contain information not contained in MC questions, (ii) this information is correlated with a measure of student knowledge and understanding of course material, and (iii) CR questions are better able to ‛explain′ academic achievement in other courses than additional MC questions. There is some evidence to suggest that this greater explanatory power has to do with the ability of CR questions to measure higher-level learning as characterised by Bloom′s taxonomy (Bloom, 1956). Both (i) the generalisability of our results to other principles of economics classes, and (ii) the practical significance (in terms of students′ grades) of our findings, remain to be determined.