Targeted Encouragement and the College Major Gender Gap
How to cite
Ramirez-Espinoza, Fernanda. 2021. "Targeted Encouragement and the College Major Gender Gap." AEA RCT Registry. AEARCTR-0007229. https://doi.org/10.1257/rct.7229-1.0
Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.
* Primary investigator
Abstract
Although educational attainment gender gaps have reversed in most high-income countries and Latin America (Goldin, 2002; Goldin, Katz, & Kuziemko, 2006; Duryea, Galiani, Nop o, & Piras, 2007), a high degree of occupational segregation remains: men still dominate highly profitable fields like science, technology, engineering, and mathematics (STEM) even when controlling for measured ability (Schneeweis & Zweimuller, 2012). Preferences in the field of study can be shaped by individuals' self-efficacy beliefs. This could be causing part of the STEM gender gap: holding performance constant, women have been found to be less confident about their own ability in math and science than men (Bordalo, Coffman, Gennaioli, & Shleifer, 2019). The hypothesis of this study is that increasing students' confidence in their ability can have effects on their college application and enrollment decisions and that this effect can be different for men than for women.
In the intervention used in this study (implemented by the Chilean Ministry of Education), top-performing students in STEM (math + science) and Humanities (language + history) tests were randomly assigned to get an encouragement message that congratulated them for their high scores and asked if they have considered STEM/Humanities majors. Students that excelled in the four tests received an encouraging message that congratulated them and mentions that with such high scores many options are available, without specifying a particular subject area. The experiment allows me to test how encouragement affects students' college enrollment decisions. By measuring gender differences in this effect, I will expand the literature that studies how students make their major choice, testing an intervention that can potentially improve enrollment. Furthermore, I will contribute to expanding the literature on women and STEM by proposing an intervention that could help close the STEM gap in college major choice.
Intervention
In this intervention, the Ministry of Education sent an information card to students via email at the time they got their University Admission Test results. A randomized sample of the top-performing students in STEM (math + science) and Humanities (language + history) tests got an encouragement message that congratulates them for their high scores and asks if they have considered STEM/Humanities majors. Students that excelled in the four tests received an encouraging message that congratulated them and mentions that with such high scores many options are available, without specifying a particular subject area.
Experimental design
The set of participants that were eligible to get the encouragement message were all students that took the university admission test in December 2020 and performed in the top 30- percent of STEM tests, humanities tests, or all four tests. To get the final samples, percentiles of STEM, humanities, and all tests within a students’ region 1 are calculated. If a student took all the tests and their average total score was in the top 30 percent for their region, they were assigned to the “All” group. All the students that were not in the “All” group but scored in the top 30 percent in either the STEM or humanities tests for their region are assigned to the “STEM” and “Humanities” groups. If students do not belong to the three previous groups, they belong to the “Other” group.
No students in the “Other” group were assigned to treatment. For the remaining three groups (All, STEM, and Humanities), fifty percent of eligible students are assigned to the treatment within each group. The remaining fifty percent was assigned to the control group
Design details
Cited by
No citation data available for this registration yet.