Sign In

Communications of the ACM

ACM TechNews

Dartmouth Researchers Create First Smartphone App That Predicts Gpa

View as: Print Mobile App Share:
The term starts with students being very active. As the term gets underway things change. Activity drops sharply to its lowest level during midterm and stays that way for the rest of the term. Sleep follows a similar pattern -- drops to a low point during

Researchers at Dartmouth College have developed a smartphone app that can predict college students' grade point averages based on smartphone data.

Credit: Dartmouth College

Dartmouth College researchers have developed SmartGPA, a smartphone app that automatically predicts college students' grade point averages based on their smartphone data, taking into account their habits on studying, partying, sleeping, exercising, and other conscious and unconscious behaviors.

The SmartGPA app is based on an earlier StudentLife study, which resulted in the first smartphone app that automatically reveals college students' mental health, academic performance, and behavioral trends.

"Our SmartGPA results show there are a number of important study and social behaviors automatically inferred from smartphone sensing data that significantly correlate with term and cumulative GPA," says Dartmouth professor Andrew Campbell.

The app relies on automatic sensing data and machine-learning algorithms to infer users' high-level behaviors. The researchers tested the app by installing it on the smartphones of 30 Dartmouth students and monitoring them over a 10-week period. The results show the app, along with periodic self-reports from students, can predict the user's GPA within 17 hundredths of a point against their cumulative GPA from their transcripts.

The findings show that high performers spent more time studying, were more conscious about their behavior, and had more instances of positive mood at the end of the term.

From Dartmouth College
View Full Article


Abstracts Copyright © 2015 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account