Competition and Pre-College Effort Dynamics

How to cite

Ozer, Esma. 2024. "Competition and Pre-College Effort Dynamics." AEA RCT Registry. AEARCTR-0015023. https://doi.org/10.1257/rct.15023-1.1

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

Esma Ozer*

* Primary investigator

Registry ID
AEARCTR-0015023
DOI
https://doi.org/10.1257/rct.15023-1.1
First registered
Last updated
Status
On going
Trial dates
2024-11-26 – 2025-12-30
Countries
Turkey ()
JEL codes
D44, C72, I20, I28, C91, C92, C93
Registry record
http://www.socialscienceregistry.org/trials/15023

Abstract

The goal of the field experiment is to create a data-generating process to structurally identify students' decision-making processes, which include preferences, production, and cost functions modeled as a function of peer learning in a competitive environment. Thus, the ultimate goal is to provide counterfactual simulations for an optimal design of competitive forces interacting with peer matching and learning. The data-generating process is inspired by Cotton et al. (2024) in that I use a website tool that includes a novel feature allowing student collaboration. Students are asked to use this website during an interim period, starting with a baseline exam and survey, followed by an endline exam and survey. The website usage allows data collection on effort choices, both individual and collaborative, in response to the competitive forces imposed by the prize structure in the final exam. I aim to use baseline and endline exams, as well as student-reported academic backgrounds, to estimate the model where the input comes from the website data, including the number of quizzes solved, daily time spent, and the number of chat messages exchanged within a pair.

Intervention

I have a 3x2 intervention design. The first dimension covers the different prize structures, and the second dimension covers whether a person is assigned to individual study mode or the peer study mode for website studying. The first arm is no competition, and the payment is based on the score in the final exam. The second is a rank-based prize among two people, and the third is a rank-based prize among ten students. The aim is to externally vary the marginal benefit to identify the model components, including production and cost functions. Separately assigning students to individual and group study modes allows estimation of the impact of peer learning.

Experimental design

I have 12 high schools with a total sample population of approximately 1,500 students. In each school, the target population consists of 10th graders. This grade was chosen because they are mature enough to be aware of upcoming college entrance exams but not as preoccupied with exams as 11th and 12th graders, who might have less time for program engagement.

First, every student takes the same survey and baseline exam. The results of the baseline exam will be used to classify students into three performance groups: Low, Medium, and High. Within each group, students are randomly assigned to one of three competition arms:

No Competition: Prizes are awarded based solely on the final exam score.

Competition 1: A rank-based prize system where one prize is awarded among two students.

Competition 2: A rank-based prize system where three prizes are awarded among ten students.

Within each competition arm, students are further assigned to either individual study mode or peer study mode. The number of students assigned to the peer study mode is double that of the individual study mode.

Students are then informed about their assigned competition arm and their standing within their performance group based on the baseline exam. Those in the peer study mode under Competition 1 and 2 also receive information about their peers’ scores and ranks within the group.

Design details

Randomization method
randomization done in office by a computer.
Randomization unit
The randomization unit is at the individual level. Students are randomized into competition arms and study modes within Low, Medium, High strata.
Clusters
Planned number of student observations is 1500.
Observations
Planned number of student observations is 1500.
Arms
For each competition group, 175 students are assigned to the individual study mode, and 350 students are assigned to the peer study mode.
Primary outcomes
The outcomes come from three main sources: baseline and endline surveys, baseline and endline exams administered by the research team, and website activity data. The survey responses will be used to identify student preferences and understand the mechanisms behind their effort choices. The baseline and endline exams will capture students' learning outcomes and proficiency gains. The website activity data provides critical effort-related information, making it one of the most important outcomes of this experiment. Its high-frequency data collection enables tracking of both competitive and cooperative behaviors, offering insights into how students respond to different incentive structures and study environments.

Cited by

No citation data available for this registration yet.

Keywords