What are we doing?
We bring together social scientists and computer scientists, using our economics of education knowledge, causal inference spirits, policy-relevant minds, and big data methods, to explore effective approaches to improve low-income students' college access and choices.
Our first five-year plan:
- Year 1 (2016): Examine what works in centralized admissions in China
- Year 2 (2017): Scale up 1-on-1 advising using incentive policies and machine learning algorithms
- Year 3 (2018): Increase program take-up for the online system
- Year 4 (2019): Artificial intelligence
- Year 5 (2020): Expand to all low-income students in China
Why?
There is rapidly growing literature on the significant impacts of information provision and individualized assistance on increasing college opportunity for low-income students the United States. But little is known about the effectiveness of these interventions in other contexts.
We aim to provide novel experimental evidence on what works in a centralized admissions system in China, and provide informational interventions and assistance to improve poor students' college opportunities.
Research projects in progress
- Increasing take-up using behavioral interventions, with Lindsay Page (Pittsburgh) and Avi Feller (Berkeley)
- Evaluation of the randomized interventions, with Yanqing Ding (Peking)
- Identifying strategies in college choice, with Shaoda Wang (Berkeley), Jiaxuan Li (Amazon)
- Retaking college entrance exam, with Di Xu (UC Irvine)
- Evaluating the free teacher college policy, with Yang Song (Colgate)
- Ranking Chinese colleges using value-added and text-as-data, with grad students at Peking
Updates
- This project website is still under construction (ingnore the tab links). Free free to email yxy@umich.edu for anything of your interest. Several computer scientists will be working on the website so that the updates are coming.
- In the 18-19 academic year, we are expanding the interventions to a large number of high school graduates through various approaches to increase take-up and the automatic online system.
- In the 17-18 academic year, we were building the Chinese version College Scoreboard. We also published a Chinese college ranking data system to aid students in college choice. Given the delayed development of the online system, we were not able to conduct large-scale randomized experiments (we conducted a few small scale ones). We provided online consultancy to moret than 30,000 studens across the country.
- Recent media mentions: China Science Daily, The CPPCC Newspaper, People.cn.
- In the 16-17 academic year, we implemented randomized experiments in three low-income provinces of China: Anhui, Ningxia, Yunnan. By August, we have received many successful admissions results from our treated kids.
- In the 15-16 academic year, we conducted our pilot RCT in Ningxia Province.
Team photos
(Kunming, June 2017)
(Sanya, Jan 2018)
Moments
(A school workshop with more than 1,000 students, Chongqing, 2018)
("I don't know how many poor students are fortunate enough to command a county mayor to file their college applications." - my WeChat post on 2017.06.25 10:30pm after working with Yinduo for our 1-on-1 advising)
("Teacher, your assistance is god's grace for students like me. I trust you." - from our treated student during the 1-on-1 advising)
Life is short. Do something right. × 2018