Researchers at Dartmouth College have successfully developed and tested an Android app which harvests live logging data to gauge a student’s stress levels and its effect on academic performance and behavior. The Android app titled StudentLife was developed by a team of researchers led by Professor Andrew Campbell. In the testing phase, 48 students at the college gave their consent to let the app collect their smartphone data for a 10-week term. The StudentLife application collected the smartphone’s location and motion data along with timings of voice and text interactions. The data helped the researchers know in which locations the students were spending their time and about the quality of social interactions. The app would also activate the microphone of the smartphone to record conversations of students from time to time. With the help of computational method and machine learning algorithms the app was able to combine data from well-known mental health surveys and mined smartphone data and was able to tell if the students were stressed and why. In the next paragraph we have highlighted some of the findings of their reports which were presented at the ACM International Joint Conference on Pervasive and Ubiquitous Computing.
The researchers found out that students who sleep more and have more social interactions were less likely to be stressed, had better grades and were less physically active. Students who were physically active were less likely to be depressed. Being lonely was an indicator of depression. Serial class bunkers will love to know that the researchers found that there was no relation between class attendance and academic performance. So next time your professor scolds you for low attendance present him/her a PDF file of this research.
The team at Dartmouth was well aware of the privacy concerns of this app. The mined data was completely anonymous and the participating students were never given the results of the analysis. To know more about the app you can head over to the project report [PDF file], Science Daily and MIT Technology Review.
0 Comments:
Post a Comment