The Impact of Question Format in Principle of Economics Classes: Evidence from New Zealand

This study investigates whether question format disadvantages certain types of students. I use assessment data compiled from principles of economics classes at the University of Canterbury from 2002-2008. I combine these with administrative data on student characteristics to create a comprehensive dataset of over 20,000 observations. To control for student ability, I use a battery of measures of student performance in non-economics classes. In the absence of controls for student ability, I find that question format appears to have a significant impact on student performance. These mostly disappear when student ability variables are added. The major exception are student characteristics associated with language: I find that non-native English speakers are relatively disadvantaged by constructed response questions even after controlling for student ability.