Smartphone app can predict students' grade point average

Developed by researchers at Dartmouth College and the University of Texas at Austin, the app in question, smartGPA is accurate to within 0.17 of a point.

The app's data - gathered via monitoring all smartphone use from physical activity to time spent sleeping - is analysed via machine learning algorithms and once periodic self reporting from users is factored in, is unnervingly accurate in predicting grades.

The app and its supporting research underline that there are certain behavioural patterns that have a direct impact on a student's grade point average - such as stress levels, time devoted to social interaction and sleep cycles - and these behaviours can be quantified via a smartphone, automatically, without need for direct user input.

The good news is that if the app recognises behaviour that could negatively impact on the student's average, it can offer alerts and advice on how to turn things around.