Please fill out this short user survey of only 3 questions in order to help us improve the site. We appreciate your feedback!
Gender Mentor Effects in University
Last registered on August 18, 2020


Trial Information
General Information
Gender Mentor Effects in University
Initial registration date
August 26, 2019
Last updated
August 18, 2020 7:06 AM EDT
Primary Investigator
University of Zurich
Other Primary Investigator(s)
Additional Trial Information
Start date
End date
Secondary IDs
This study examines the impacts of mentors, especially their gender and gender-related characteristics, on college students' educational expectation, performance and persistence. First-semester bachelor students can voluntarily register for a mentoring program. If registered, they will be randomly assigned to an upper-year student mentor from the same study program. Thanks to the exogenous variation of mentors across groups, I can study the causal mentor effects.
External Link(s)
Registration Citation
Shan, Xiaoyue. 2020. "Gender Mentor Effects in University." AEA RCT Registry. August 18. https://doi.org/10.1257/rct.4599-2.0.
Experimental Details
Intervention Start Date
Intervention End Date
Primary Outcomes
Primary Outcomes (end points)
(1) Expectations and beliefs about education outcomes;
(2) Performance in first-semester courses and later courses;
(3) Dropout, major choice and degree completion.
Primary Outcomes (explanation)
Educational expectations and beliefs will be measured in the student survey. Educational outcomes in the short run and long run will be acquired from university administrative data.
Secondary Outcomes
Secondary Outcomes (end points)
Study habits (e.g., study hours); beliefs about ability and gender roles; self-esteem.
Secondary Outcomes (explanation)
These outcomes will be measured in the student endline survey.
Experimental Design
Experimental Design
The experiment will randomly assign mentees (first-semester students) to mentors (second-/third- year students). Mentors help mentees to adapt to university life, by sharing information and organizing social/academic events.
Experimental Design Details
The mentoring program is organized by the university. On the mentor's side, students from the second year or third year are recruited to work as mentors. On the mentee's side, all new students majoring economics, business and computer science can sign up for a mentor. Then mentees are randomly assigned to mentors from the same study program (each mentor has approximately 15 mentees). In the beginning of the fall semester 2019, both mentees and mentors are invited to fill out a survey, which measures their demographic backgrounds, educational expectations and beliefs, gender attitudes, socioeconomic preferences, and personality traits. During the semester, mentors send newsletters and organize social/academic events, to help mentees get adapted to university life. In the end of the semester, another survey will be conducted for new students. The survey measures students' study habits, social network, and their beliefs about educational outcomes, gender roles and own ability.
Randomization Method
The randomization (match between mentors and mentees) will be conducted by the principal investigator using STATA in office, given the list of mentors and mentees.
Randomization Unit
The unit of randomization is at the individual level, and stratified at the level of study programs. Mentees are assigned to mentors from the same study program.
Was the treatment clustered?
Experiment Characteristics
Sample size: planned number of clusters
There will be approximately 35 mentors, and each mentor corresponds to a group of approximately 15 mentees.
Sample size: planned number of observations
The exact number of observations (mentees) depends on the sign-up rate. An educated approximation is 525.
Sample size (or number of clusters) by treatment arms
When examining the impact of mentors' gender, the two treatment arms are (1) students with female mentors and (2) students with male mentors. The relative sample size depends on the gender composition of recruited mentors. For instance, if there are 10 female mentors out of 35, the female-mentor treatment arm has approximately 150 students and the male-mentor treatment arm has approximately 375 students. When examining the impact of mentors' gender attitudes (a continuous measure), there's no clear differentiation of treatment arms.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Based on 2018 student data, I use Monte Carlo simulations to conduct power calculation. Assuming that 525 students are randomly assigned to 35 mentors (11 female and 24 male), with 10% significance level and 80% power, the experiment can detect a gender mentor effect equal to 11% standard deviation of standardized scores in a compulsory first-semester course. With 5% significance level and 80% power, the experiment can detect a gender mentor effect equal to 13% standard deviation of standardized scores. The gender mentor effect is defined as the impact of a female mentor on student outcomes, controlling for student baseline characteristics (e.g., gender, age, major, personality traits).
IRB Name
University of Zurich
IRB Approval Date
IRB Approval Number
OEC IRB # 2019-033
Post Trial Information
Study Withdrawal
Is the intervention completed?
Is data collection complete?
Data Publication
Data Publication
Is public data available?
Program Files
Program Files
Reports, Papers & Other Materials
Relevant Paper(s)