Drop out From School in Mexico: An Empirical Analysis With Mixed Logit Model


Engineering and Applied Sciences


Professor of Economics

Project Summary

The goal of my project was to build a probability model that predicts a Mexican teenager’s dropout rate based on his or her socioeconomic status. I got a chance to practice my computer science and econometrics knowledge in real research experience.

The key difference between a research and course project is that I have to myself gather all the needed information and programming techniques either by reading related papers in the research field or requesting data from a nonprofit organization. At the beginning of my project, I had to explain my research plan and goal to the University of Minnesota to get access to its database of Mexico census from 2010. In the following weeks, I intensively practiced analyzing census data with the statistical programming language R and summarized results using fact sheets, graphs, and tables. The coolest part was I generated a web scraping program to extract information from over 130,000 schools in Mexico. Extracting information such as location coordinates and school type, I was able to compile them into a database that is not only useful to my project but also beneficial to other research projects.

During the ten weeks of PURM, I also realized how important teamwork is in research. My colleague Beverlye helped me translated Mexican government documents because I had no idea about Spanish. Sometimes I had to confront problems that were out of knowledge, and I was grateful my professor Petra Todd and graduate assistant Gabrielle Vasey provided me with insightful advice for solutions and new directions.

Throughout this research experience, I had the opportunity to learn about the Mexican education system, geography, economics, culture, etc. The practice of knowledge learned in class was also extremely rewarding and I would like to continue my future research in econometrics.