Timing Is Everything: Experimental Evidence on Quiz Timing and Exam Performance in Introductory Economics

How to cite

Fredrickson, Amanda. 2026. "Timing Is Everything: Experimental Evidence on Quiz Timing and Exam Performance in Introductory Economics." AEA RCT Registry. AEARCTR-0019210. https://doi.org/10.1257/rct.19210-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Amanda Fredrickson*

* Primary investigator

Registry ID
AEARCTR-0019210
DOI
https://doi.org/10.1257/rct.19210-1.0
First registered
Last updated
Status
In development
Trial dates
2026-08-17 – 2027-05-28
Countries
Private
JEL codes
A22 (Economic Education and Teaching of Economics)
Registry record
http://www.socialscienceregistry.org/trials/19210

Abstract

Short, low-stakes quizzes are a common teaching tool in college courses, but little is known about whether the timing of those quizzes matters for student learning. This study examines whether administering a five-minute quiz at the end of class, after instruction, produces better exam performance than administering the same quiz before class. We conduct a randomized controlled trial in SS201: Principles of Economics at the United States Military Academy (West Point), randomly assigning course sections to one of two quiz-timing formats. All other aspects of the course, content, assignments, exams, and grading, are identical across conditions. The study runs across two consecutive semesters (Fall 2026 and Spring 2027), with the primary outcome being performance on the standardized Term-End Exam. If confirmed, findings would suggest a simple, costless improvement to quiz design that any instructor could implement without changing course content or grading burden.

Intervention

Sections are randomly assigned to one of two quiz-timing conditions: (1) End-of-Class — cadets complete a 5-question multiple-choice quiz in the final 5–10 minutes of the lesson period, after instruction has been delivered; or (2) Pre-Class — cadets complete the same 5-question quiz before arriving to class. Quiz content is identical across conditions within any lesson. All other course elements — syllabus, WPRs, TEE, grading rubrics, and instructional contact hours — are held constant. Each instructor teaches both conditions simultaneously across their assigned sections.

Experimental design

Within-instructor RCT. Sections are randomly assigned to end-of-class or pre-class conditions prior to each semester, stratified by instructor and time of day (morning vs. afternoon). Each instructor with four sections teaches two end-of-class and two pre-class sections; each instructor with two sections teaches one of each. Single-section instructors are assigned to end-of-class by default and excluded from the primary analysis. The primary analysis uses OLS regression with instructor and time-of-day fixed effects, standard errors clustered at the section level.

Design details

Randomization method
R statistical software (set.seed(2026201)), stratified by instructor and time of day
Randomization unit
Section (cluster)
Clusters
48 sections (24 per semester)
Observations
864 (approximately 432 per semester)
Arms
Approximately 432 end-of-class, 432 pre-class (pooled across both semesters)
Primary outcomes
Term-End Exam (TEE) score (250-point scale)
Secondary outcomes
WPR (midterm) scores; quiz–TEE partial correlations by condition; quiz completion rates; subgroup TEE performance by CEER quartile, sex, recruited athlete status, prior service status, and USMAPS attendance

Cited by

No citation data available for this registration yet.

Keywords