This paper uses a field experiment conducted in high schools from the Los Angeles Unified School District to understand the nature of social costs students face when they make decisions on educational investments.
External Link(s)
Citation
Bursztyn, Leonardo and Robert Jensen. 2014. "Understanding the role of peer pressure in shaping students educational investments." AEA RCT Registry. February 12. https://doi.org/10.1257/rct.60-1.0.
Take up rates of the offer of free SAT prep courses.
Primary Outcomes (explanation)
Secondary Outcomes (end points)
Secondary Outcomes (explanation)
Experimental Design
We will provide students with forms that ask them whether they would like to sign up for a free, online SAT prep course.
Experimental Design Details
We will provide students with forms that ask them whether they would like to sign up for a free, online SAT prep course. We will randomly vary (among students within classrooms) whether the form states that the information will be kept completely private from everyone, including their classmates. The forms will be identical other than this final disclaimer. The more explicit, greater assurance of privacy from classmates regarding the decision to enroll in the test prep course is proposed to reduce the level of social stigma involved in enrollment and thus increase take up (note however, in none of the cases will we make sign up decisions public--we simply vary the extent to which privacy from classmates is emphasized).
We will also stratify by honors vs. non-honors classes. Our model suggests that different norms may take hold in the two kinds of classes, with stigma greater in non-honors classes, thus a greater effect of information being kept private.
We will also collect data on self-reported attributes such as grades and time preferences, where we expect students who do better to face less of a stigma, students who discount the future more to be more concerned with stigma, as will students who care more about popularity.
Randomization Method
The differing sign up forms will be shuffled and distributed to students in the order they sit in the room.
Randomization Unit
Students, within classrooms (and stratified for honors vs. non-honors classes)
Was the treatment clustered?
Yes
Sample size: planned number of clusters
4 schools, approximately 20 classrooms
Sample size: planned number of observations
1,000 (approximate, since classroom sizes will vary)
Sample size (or number of clusters) by treatment arms
Approximately 500 treatment, 500 control (approximate since classroom sizes will vary)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)