Experimental Design
The experiment will be conducted in 18 middle schools in a local county in China, with more than one third of the students suffering from parental absence (being "left behind"). In each school, two cohorts are invited to participate in our study: students in grade 7 (who entered middle school in September 2022) and grade 8 (who entered middle school in September 2021).
Our experiment involves one treatment and one control. To avoid imbalances in treatment assignment and increase precision of the estimates, we employ stratified randomization at the class level. The classes will be stratified by school year and the number of students that are left behind (below/above median).
The timeline of our intervention is as follows. We initiate student recruitment two weeks prior to the intervention to allow sufficient time for head teachers to secure consent from students and their parents. Concurrently, the recruitment of instructors will also be launched in collaboration with the local Education Bureau. The entire recruitment process is expected to be completed within a two-week timeframe. One weeks ahead the intervention, our research team will visit each school to run the baseline survey. Training sessions are scheduled on that week.
Our intervention period spans 6 weeks, during which the students in the treated classes will participate in a well-designed curriculum centered around the growth mindset. Through compelling scientific evidence, lively videos, mini case studies, and other in-class activities, we emphasize the benefits of cultivating a growth mindset and provide practical tools to operationalize this mode of thinking. Additionally, we will assign after-class assignments to provide students with opportunities to apply what they have learned. Meanwhile, students in the control group will continue attending their regular class meetings as usual.
To avoid the direct priming from course material, our endline survey and supplementary experiment will be carried out one week after the intervention.
Since our interest lies in both the short term and medium-term effect of the intervention, we will track the academic performance of these students in the following semesters, utilizing the school administrative data.