Professor will this be on the Exam?

How to cite

Joyce, Theodore; Dench, Daniel. 2018. "Professor will this be on the Exam?." AEA RCT Registry. AEARCTR-0002669. https://doi.org/10.1257/rct.2669-2.0

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

Theodore Joyce* , Daniel Dench

* Primary investigator

Registry ID
AEARCTR-0002669
DOI
https://doi.org/10.1257/rct.2669-2.0
First registered
Last updated
Status
Completed
Trial dates
2017-12-27 – 2018-08-30
Countries
United States of America (Northeast)
JEL codes
I21
Registry record
http://www.socialscienceregistry.org/trials/2669

Abstract

Today's students face many pressures vying to capture their attention, and undergraduate curricula increasingly rely on self-directed content exposure that occurs outside of the traditional classroom. This combination may exacerbate behavioral failures inhibiting human capital production. In a previous experiment we found that email nudges to focus on certain ungraded problems available online had little impact on student performance on questions very similar to those on the exam. We realized that students focused on the graded assignments with relatively few attempting to do the ungraded “nudged” problems. In this study we test whether nudges to focus on graded assignments with the message that such problems are likely to appear on the exam sent to a random subset of students within each class help students focus on core concepts more than when graded assignments are not nudged. In addition, we test whether nudges to practice ungraded problems also with the hint that “questions like these are likely to appear on the exams” increases student time spent online with the ungraded problems and enhances performance on the test. Nudges to focus on practice problems will also be sent to a random sub-sample within each class. Lastly we will vary the amount each graded assignment counts towards their final grade to see if exercises that count more garner more online effort and improve performance on exams.

Experimental design

The study design will use within- classroom variation in graded problems to identify the effect of online practice problems in teaching basic economic concepts to students. The study will also use within- person variation in the percent towards the final grade that each question is worth from week to week.

Design details

Randomization method
Done by program in stata by classroom using set seed and runiform() packages.
Randomization unit
Individual/Student
Clusters
0 clusters
Observations
Approximately 660 students, although this may depend on registration and course drop patterns after the semester begins. In some estimations we will use two observations per student (for midterm and final exams) or estimate directly on question-level outcomes (as detailed in the full analysis plan).
Arms
Approximately 330 individuals per treatment arm, with 2 treatment arms (A,B)
Primary outcomes
Probability that a problem set was completed (binary), correct answer to questions corresponding to intervention problems (binary), combined score on questions corresponding to intervention problems (integer), Activity access corresponding to intervention (binary), combined activity access corresponding to intervention (integer), total time accessing activities related to intervention (log).

Cited by

No citation data available for this registration yet.

Keywords