Information and Academic-Track Choice in Rural China: A Field Experiment

How to cite

Peng, Mengying; Li, Xin; Ye, Xiaoyang. 2026. "Information and Academic-Track Choice in Rural China: A Field Experiment." AEA RCT Registry. AEARCTR-0019231. https://doi.org/10.1257/rct.19231-1.0

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

Mengying Peng* , Xin Li , Xiaoyang Ye

* Primary investigator

Registry ID
AEARCTR-0019231
DOI
https://doi.org/10.1257/rct.19231-1.0
First registered
Last updated
Status
Completed
Trial dates
2024-11-18 – 2024-12-31
Countries
China (Fujian)
JEL codes
D83, I23, I24, I25, J24, O15
Registry record
http://www.socialscienceregistry.org/trials/19231

Abstract

This study examines whether providing information can help students make more effective subject-portfolio choices under China's New Gaokao reform, which replaces two broad tracks with a larger set of subject combinations tied to college-major eligibility. In a field experiment with high-school students in a rural county in Fujian Province, students were randomly assigned to receive one of several types of information about their subject-portfolio option - covering admission coverage, academic difficulty and major eligibility, or expected wages—or to a status-quo academic-planning condition. The study measures how information affects students' stated subject-portfolio preferences, including whether they revise their choices, which dimensions of the choice they adjust, and the internal consistency of their stated preferences. The setting speaks to a broader question of how resource-constrained education systems can support students in navigating increasingly complex educational choices.

Intervention

Students were randomly assigned to receive one of three types of information about their subject-portfolio options under the New Gaokao, or to a status-quo academic-planning control. The three information arms provided, respectively: (1) admission-seat coverage, showing how many college-major options each portfolio keeps open; (2) academic difficulty and eligibility information; and (3) expected-wage information associated with different portfolios. Information was delivered within a survey instrument. Students' preferred subject portfolios were elicited before and after the information.

Experimental design

The study is a four-arm survey experiment conducted with high-school students in a rural county in Fujian Province, China. Students' baseline subject-portfolio preferences were elicited, after which they were randomly assigned to one of three information conditions (admission-seat coverage; academic difficulty and eligibility; expected wages) or a status-quo academic-planning control. Post-information preferences were then elicited. The experiment was conducted separately with a Grade 10 cohort (whose portfolio choice remained flexible) and a Grade 11 cohort (whose portfolio had already been declared).

Design details

Randomization method
Assignment performed by the survey platform built-in randomizer.
Randomization unit
Individual student.
Clusters
Not clustered
Observations
4,082 students (2,714 Grade 10 and 1,368 Grade 11) in the analytical sample.
Arms
Grade 10: 683 control, 687 admission-coverage, 665 difficulty-and-eligibility, 679 wage. Grade 11: 346 control, 350 admission-coverage, 334 difficulty-and-eligibility, 338 wage. Combined: 1,029 control, 1,037 admission-coverage, 999 difficulty-and-eligibility, 1,017 wage (total 4,082).
Primary outcomes
The primary outcome is whether a student changed their stated subject-portfolio preference after receiving information (an indicator for any change between the pre- and post-information elicitation). Related primary outcomes include whether the change involved a within-track elective swap versus a track switch.
Secondary outcomes
Secondary outcomes include preference formation among initially undecided students (whether a student with no baseline preference articulates one), the internal coherence of stated preferences, and heterogeneity of treatment effects by student academic performance and socioeconomic background.

Cited by

No citation data available for this registration yet.

Keywords