Testing Inequality: Skill building on Standardized Exam Outcomes

How to cite

SEGOVIA, JOSELIN; Bettinger, Eric; Urgilés, María Paz. 2025. "Testing Inequality: Skill building on Standardized Exam Outcomes." AEA RCT Registry. AEARCTR-0016097. https://doi.org/10.1257/rct.16097-1.0

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

JOSELIN SEGOVIA* , Eric Bettinger , María Paz Urgilés

* Primary investigator

Registry ID
AEARCTR-0016097
DOI
https://doi.org/10.1257/rct.16097-1.0
First registered
Last updated
Status
Completed
Trial dates
2022-09-01 – 2025-02-28
Countries
Ecuador (South America)
JEL codes
C93, I24, J24, 015
Registry record
http://www.socialscienceregistry.org/trials/16097

Abstract

College enrollment remains highly stratified by socioeconomic status throughout the world and especially in Latin America. College entrance exams have received worldwide criticism about its validity in countries such as the US, China, Brazil, Turkey and Chile, with a shared concern for the role of wealth in student outcomes. The importance of college exams has led to a significant market for test preparation services, which is most likely accessible to economically advantaged students. We consider a simple case where preparation can lead to two types of knowledge are necessary for entrance exams – cognitive, measurable knowledge and test-specific knowledge, knowledge about the process and rules of the exam. We focus on providing test taking preparation to economically disadvantaged students. We show that socioeconomic gaps can be reduced by augmenting both types of knowledge. The training focused on both mathematical knowledge and detailed information about the nature of the exam.

Experimental design

This study uses a randomized controlled trial (RCT) to evaluate the impact of providing economically disadvantaged students with preparation for standardized tests. The study targets students entering third year of public high schools in the urban area of Cuenca, Ecuador. We randomly assign classrooms within schools to the treatment group, which received training that covered topics on numerical and logical reasoning through both guided and independent practice. The control group did not receive any coaching and continued with the regular school activities. The schools are selected using a national performance ranking from past entrance exams to the public university system. The program is implemented within school class hours in coordination with the directors in the participating schools. The program evaluation compares outcomes such as scores, attempts and correctedness between students in treatment and control schools to assess the effectiveness of the intervention.

Design details

Randomization method
Randomization done in office
Randomization unit
Randomization unit is the classroom, with measurement at the individual level
Clusters
The number of clusters is 75 classrooms
Observations
The full sample consists of 1,955 students
Arms
Around 975 students in the treatment and 975 students in the control group
Primary outcomes
The primary outcome is the after-treatment score achieved, the number of correct answers and the number of attempted answers.

Cited by

No citation data available for this registration yet.

Keywords