Student Dropout Prediction in Regional Universities Using Automated Machine Learning

Bin Chen

Download Paper | Permalink

Conference: 2024 ASEE North Central Section Conference
Location: Kalamazoo, Michigan
Publication Date: March 22, 2024
Start Date: March 22, 2024
End Date: March 23, 2024
Page Count: 7
DOI: 10.18260/1-2--45636
Permanent URL: https://peer.asee.org/45636
Download Count: 20

Request a correction

Paper Authors

author page

Bin Chen Purdue University Fort Wayne

Download Paper | Permalink

Abstract

The overall dropout rate of engineering students is between 40% to 50% in the United States, according to the American Society for Engineering Education (ASEE) survey from 2009 to 2018. The severity of dropout is not the same among all universities. Regional universities have experienced much higher rates of student attrition from engineering programs compared to Carnegie R1 universities. Reducing the dropout rate in regional universities becomes the most effective and economical way to increase retention and graduation for the nation. However, most studies are based on nationwide data or data from national universities, rather than regional universities. Students in regional universities are more diverse in personal background and academic preparation. They often face unique challenges, such as financial constraints, daily commutes, and the need to balance academic pursuits with employment and family responsibilities. Focusing on a few statistically significant factors and applying resources to improve them is not the most effective way to reduce dropout rates.

This study is to develop a machine learning framework with the capability of learning from heterogeneous data to identify students at risk of dropout from five main data sources:

1) high school information, 2) demographic information, 3) college and department program information, 4) academic information, course study and research activities, 5) student real time feedback to the web, mobile phone apps and course learning management systems

The first three categories change less frequently and have stable and long-term effects on the decision of dropout. The data in categories 4 and 5 have intermediate to high variations. The combination of all above data will include long-term to short-term influences on dropout decisions in a static, dynamic, and cumulative manner.

The data will be preprocessed and then split into training, validation and testing datasets respectively for machine learning. The predicted risk of a student dropping out from an engineering program will be a probability between 100% for graduation and 0% for dropout. The completion of applied credits for a degree program will be linearly scaled between 0 and 1 as targets for supervised learning. If a student dropped out of the engineering program the following semester, the probability of graduation would be 0, otherwise, a new graduation probability will be assigned to the student until that student either graduates or drops out. Students will be ranked based on their predicted dropout probability, and the group with the highest dropout probabilities will be informed of their eligibility to enroll in services, subject to the students' own volition.

Regional universities often serve as gateways to higher education for underrepresented groups, including first-generation college students, individuals with disabilities, and minorities. By understanding and addressing these challenges with early identification of at-risk students, regional universities can play a crucial role in increasing the representation of underrepresented groups in STEM fields.

Citation
Format

Chen, B. (2024, March), Student Dropout Prediction in Regional Universities Using Automated Machine Learning Paper presented at 2024 ASEE North Central Section Conference, Kalamazoo, Michigan. 10.18260/1-2--45636

TY - CPAPER
AB - The overall dropout rate of engineering students is between 40% to 50% in the United States, according to the American Society for Engineering Education (ASEE) survey from 2009 to 2018. The severity of dropout is not the same among all universities. Regional universities have experienced much higher rates of student attrition from engineering programs compared to Carnegie R1 universities. Reducing the dropout rate in regional universities becomes the most effective and economical way to increase retention and graduation for the nation. However, most studies are based on nationwide data or data from national universities, rather than regional universities. Students in regional universities are more diverse in personal background and academic preparation. They often face unique challenges, such as financial constraints, daily commutes, and the need to balance academic pursuits with employment and family responsibilities. Focusing on a few statistically significant factors and applying resources to improve them is not the most effective way to reduce dropout rates.

This study is to develop a machine learning framework with the capability of learning from heterogeneous data to identify students at risk of dropout from five main data sources:

1) high school information,
2) demographic information,
3) college and department program information,
4) academic information, course study and research activities,
5) student real time feedback to the web, mobile phone apps and course learning management systems

The first three categories change less frequently and have stable and long-term effects on the decision of dropout. The data in categories 4 and 5 have intermediate to high variations. The combination of all above data will include long-term to short-term influences on dropout decisions in a static, dynamic, and cumulative manner.

The data will be preprocessed and then split into training, validation and testing datasets respectively for machine learning. The predicted risk of a student dropping out from an engineering program will be a probability between 100% for graduation and 0% for dropout. The completion of applied credits for a degree program will be linearly scaled between 0 and 1 as targets for supervised learning. If a student dropped out of the engineering program the following semester, the probability of graduation would be 0, otherwise, a new graduation probability will be assigned to the student until that student either graduates or drops out. Students will be ranked based on their predicted dropout probability, and the group with the highest dropout probabilities will be informed of their eligibility to enroll in services, subject to the students&#39; own volition.

Regional universities often serve as gateways to higher education for underrepresented groups, including first-generation college students, individuals with disabilities, and minorities. By understanding and addressing these challenges with early identification of at-risk students, regional universities can play a crucial role in increasing the representation of underrepresented groups in STEM fields.
AU - Bin Chen
CY - Kalamazoo, Michigan
DA - 2024/03/22
PB - ASEE Conferences
TI - Student Dropout Prediction in Regional Universities Using Automated Machine Learning
UR - https://peer.asee.org/45636
DO - 10.18260/1-2--45636
ER -