Where data meets social impact

In 2019, with the hope to investigate social inequality by using my knowledge of applied statistics, I led a team, heading for Hejin, a rural area of China, aiming to fully evaluate the an ongoing educational policy and look for ways to promote it nationwide so as to ameliorate the uneven distribution of educational resources within China. First, we designed and distributed different questionnaires to school managers, teachers and students, so as to understand their attitudes toward this policy and how it helped them to improve. Then we visualized the data collected using R, to analyze the degree of satisfaction among different students and teachers. Second, we detected and eliminated the outliers of the historical data of the school’s graduates so as to predict the college admission rate using linear regression. Surprisingly, we found that, with a confidence level of 95%, holding other variables constant, the actual admission rate of the graduates with this policy was just 2.5 points higher than the predicted figure. What’s more, we set up an input-output model to calculate the Return on Investment (ROI). The input came from the government’s funding to support this policy. The output consisted of the discounted income cash flow of extra students who benefited from this policy and went to universities. The ROI of this policy based on the model was shockingly 400%. In our final report, we recognized the feasibility of this policy and provided some recommendations to further elevate its performance. The report was awarded with the highest honor by the school committee and the policy was carried out in many more rural areas to achieve educational equality.

Behind all the analysis and equations, what really haunts me is the students and teachers I met through out the way, those who strives to thrive and effortlessly bring hope and power to the students there. I’m also very previleged to offer my help.