Working Paper: NBER ID: w22841
Authors: Christopher Avery; Oded Gurantz; Michael Hurwitz; Jonathan Smith
Abstract: Mapping continuous raw scores from millions of Advanced Placement examinations onto the 1 to 5 integer scoring scale, we apply a regression discontinuity design to understand how students’ choice of college major is impacted by receiving a higher integer score despite similar exam performance to students who earned a lower integer score. Attaining higher scores increases the probability that a student will major in that exam subject by approximately 5 percent (0.64 percentage points), with some individual exams demonstrating increases in major choice by as much as 30 percent. These direct impacts of a higher score explain approximately 11 percent of the unconditional 64 percent (5.7 percentage points) gap in the probability of majoring in the same subject as the AP exam when attaining a 5 versus a 4. We estimate that a substantial portion of the overall effect is driven by behavioral responses to the positive signal of receiving a higher score.
Keywords: Advanced Placement; college major choice; regression discontinuity
JEL Codes: I21; I23; J24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Higher AP exam scores (5) (Y40) | Increased likelihood of majoring in the corresponding subject (Y80) |
| Higher AP exam scores (5) (Y40) | Increased likelihood of majoring in STEM fields (I23) |
| Higher AP exam scores (5) (Y40) | Higher perception of ability (D29) |
| Higher AP exam scores (5) (Y40) | Increased probability of majoring in the same subject compared to scoring 4 (C12) |