Back to History

Fields Changed

Registration

Field Before After
Trial Status in_development completed
Abstract We intend to study how a mentor program will affect academic performance and other non-cognitive traits. Volunteer mentoring for disadvantaged youth is widely recognized as beneficial for youth development. However, relatively little research has examined the effects of online mentoring programs on mentees, and the impact of mentoring experiences on mentors themselves has been largely overlooked. To address these gaps, we organized an online mentoring program that connects students from a top Chinese university with middle school students from rural areas. The study adopts a two-sided randomization design. First, college student applicants were randomly assigned either to serve as mentors for rural students or to a control group. Second, middle school applicants were simultaneously randomized to either receive mentoring or not, allowing us to assess the causal impact of the program on their academic performance. Third, mentors and mentees were randomly matched within the program. Together, this two-sided randomization and matching design enables us to examine the effects of the mentoring program on both mentors and mentees across multiple dimensions, including academic outcomes, non-cognitive abilities, and social preferences.
Last Published February 22, 2021 12:05 PM January 26, 2026 07:55 PM
Intervention (Public) Students will be either placed in the mentor program versus not. We designed an online mentoring program in which college students applied to serve as mentors and middle school students applied to participate as mentees. College student applicants were randomly assigned to either a treatment group or a control group. The treatment group served as mentors during the spring semester, while the control group did not receive the opportunity to participate during the study period. Middle school applicants were similarly randomized into treatment and control groups. The treatment group received mentoring in the spring semester, whereas the control group did not. Mentors were randomly matched with mentees on a one-to-one basis.
Primary Outcomes (End Points) Academic performance We examine distinct primary outcomes for each participant group: Mentors: The key outcome variable is social preferences. Mentees: The key outcome variable is academic performance and non-academic abilities.
Primary Outcomes (Explanation) It is evaluated by test scores Social Preferences: Elicited through economic games Academic Performance: evaluated via test scores Non-congnitive Abilities: elicted through survey questions
Experimental Design (Public) This experiment is to assign college students as mentors while middle-school students as mentees with the hope that college students can help middle-school students during communication. We organized an online mentoring program connecting college student volunteers with rural middle school students. To evaluate the program's impact on both groups, we employed a two-sided randomization design: (1) College applicants were randomly assigned to either a treatment group (serving as mentors in the Spring semester) or a control group (not offered the opportunity to participate during the study period); and (2) Middle school applicants were randomly assigned to either a treatment group (receiving mentoring) or a control group (no mentoring). Within the treatment groups, mentors and mentees were randomly paired on a one-on-one basis. The study primarily measures the impact on mentees' academic performance (via test scores), non-cognitive abilities (via survey questions), and mentors' social preferences (evaluated via economic games). Data was collected through baseline and endline surveys, as well as administrative records.
Sample size (or number of clusters) by treatment arms 600 vs 600 The study involves a total of roughly 400 college students (randomized into 200 treatment and 200 control) and 800 middle school students (randomized into 400 treatment and 400 control).
Keyword(s) Behavior, Education Behavior, Education
Secondary Outcomes (End Points) Preferences
Building on Existing Work No
Back to top