Information about Peer Altruism and Preferences for Prosocial Jobs: Evidence from Medical Students in China

How to cite

Yang, Jianan. 2026. "Information about Peer Altruism and Preferences for Prosocial Jobs: Evidence from Medical Students in China." AEA RCT Registry. AEARCTR-0018881. https://doi.org/10.1257/rct.18881-1.0

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

Jianan Yang*

* Primary investigator

Registry ID
AEARCTR-0018881
DOI
https://doi.org/10.1257/rct.18881-1.0
First registered
Last updated
Status
On going
Trial dates
2026-05-01 – 2026-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/18881

Abstract

This study estimates medical students’ preferences for job attributes and their willingness to pay (WTP) for different job characteristics as they approach the labor market. We further examine whether and how peer-comparison information about altruism causally shifts these preferences. We measure social preferences with an incentivized task and elicit job-attribute preferences through a hypothetical discrete choice experiment. We first document how beliefs about one's own social preferences relate to job preferences. We then randomly provide peer-comparison feedback to test how updating beliefs—about both the population distribution of altruism and one's own rank within it—affects WTP.

Experimental design

Our study sample consists of medical students in their graduating year who are approaching the labor market. The experiment uses a between-subjects information treatment design. When participants log in to the experiment website, they are randomly assigned to one of three groups: a control group (T1 = 0, T2 = 0; N = 400), treatment group 1 (T1 = 1, T2 = 0; N = 400), or treatment group 2 (T1 = 0, T2 = 1; N = 400). The information intervention is placed after the dictator game and prior-belief elicitation, and before the job choice experiment.

Design details

Randomization method
Randomization was conducted before the start of the experiment. We implemented blocked randomization by gender and school, and the resulting treatment assignment was preloaded into the experiment website. We assessed balance across treatment groups by regressing baseline individual characteristics on the two treatment indicators, T1 and T2. These characteristics include gender, ethnicity, parental education, parental occupation, whether any family member works in the healthcare industry, place of residence, family income, and national college entrance exam score.
Randomization unit
individual
Clusters
No cluster design
Observations
Around 1,200 medical students
Arms
Around 400 individuals in control and each of the two treatment groups
Primary outcomes
1) Altruism weight; 2) Equality-efficiency tradeoff; 3) Preferences for job attributes; 4) Willingness to pay (WTP) for job attributes.
Secondary outcomes
1) Prior beliefs about peer average altruism and own relative ranking. 2) Heterogeneous responses to peer-comparison information, including by prior beliefs. 3) Heterogeneity by gender, baseline altruism, socioeconomic background, rural/urban origin, and exposure to the medical profession. 4) Attitudes toward and experience with medical AI tools. 5) Actual job choices and perceived job attributes. 6) Final employment outcomes, conditional on successful follow-up.

Cited by

No citation data available for this registration yet.

Keywords