Skip to content

Model building to predict various measures of student success in college courses

Notifications You must be signed in to change notification settings

willsmorgan/Student-Outcomes-in-Higher-Education

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Student-Outcomes-in-Higher-Education

Model building to predict various measures of undergraduate student success

Purpose

Our goal is to test a handful of machine learning techniques in predicting outcomes in college courses. Data is gathered using the host university's Student Information System (SIS) and accounts for a wide array of student and course characteristics. Cleaning is (mostly) done in Stata 14, while the rest of the project is written in R.

The project proceeds in two directions - we first split our sample into course-level data sets and for a given model type and measure, estimate a model for each course. This allows us to extract effects unique to individual courses, at the cost of a smaller sample. The second portion of this project treats all courses in the sample as one, and estimates a model for the entire data set. This analysis is intended to be complementary to the initial course-level models, as it gives us more informed descriptions of the macro environment.

Completed scripts

  • Course-level sample selection
  • Data preparation for modeling (Model Prep.R)
  • Model Evaluation tool

In-progress scripts

  • Model definitions
  • Model estimation and evaluation
  • Population-level sample selection

About

Model building to predict various measures of student success in college courses

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published