The Problem (Do financial aid and lower total cost help increase graduation rate? ):

Only about 60% of students who begin studying for a bachelor’s degree succeed in attaining a diploma. Many institutions of higher education are facing pressure from state and federal government to increase graduation rates. Therefore, many encouragement policies for graduation rate are proposed, e.g., financial aid, reduction of study cost, etc. However, is there a relationship between these encouragement policies and graduation rate? This relationship is particularly impactful in schools with large percentages of low incomes students.

The Solution (Using time difference linear models to figure out relationship.):

We first proposed two linear models about financial aid effect for graduation rate and total study cost effect for graduation rate. There are some issues at these linear models: (1) Some crucial explanatory variables are omitted due to insufficiency of data set, e.g., type of degree, quality of education; (2) The error term is correlated with our explanatory variables, e.g., good schools cost more.

In order to resolve these issues, we consider time difference linear model for aforementioned two models by using variation of variables over a period of time as new variables. With this new model, linear regression assumptions are satisfied. This result suggests colleges and universities should not try to provide incentives or grants as methods to improve completion rates.

Besides, we also study the effect of financial aid with respect to sex for university/college enrollment.

Project Website

For more details, check out the project's website here.

Important Figures

Two Models

Problems with Both Models

Difference Model is Flat