|
Activity Number:
|
284
- GMM, Triple Joint Modeling, Bootstrapping, and Multiple Membership of Correlated Data
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 1, 2017 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
WNAR
|
|
Abstract #324241
|
|
|
Title:
|
Impact of Instructors and Majors on Student Performance Bayesian Multiple Membership Logistic Regression Model
|
|
Author(s):
|
Maria Elsa Vazquez*
|
|
Companies:
|
Arizona State University
|
|
Keywords:
|
Higher education ;
hierarchical structure ;
correlated data ;
binary outcomes
|
|
Abstract:
|
This is the third of four related papers addressing the intraclass correlation due to the hierarchical structure of the data. It is well known that consistent progress and good performance are crucial in a student's continued pursuit of a four-year degree. As such, universities have used different analytic monitoring systems to keep track of retention ratio and improve graduation rates. Such rates are key measures when recruiting and competing for the best students for the institution. We explore the hierarchical nature of the data. We present a binary multiple membership model to identify and to determine continued performance based on instructors and majors, where we consider that students are not completely nested within instructors. In addition, we looked at certain factors that significantly impact the student's performance measure. We focused on a binary measure of success initially based on dichotomized semester GPA at 3.0 then at 2.0. Multiple membership logistic regression models with Bayesian estimates were fitted to three consecutive semesters of data. We found that instructors play the most important role in student's performance.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2017 program
|
