This research aims to show that personality traits can be inferred from mobile phone usage data. The researchers analyzed data from a study of 54 participants, collecting information on their phone calls, texts, locations and survey responses over 10 weeks. They defined several metrics to quantify social activity, engagement, mobility patterns and diurnal rhythms. Regression and classification models showed that combinations of phone data features could predict individuals' risk-taking propensities with 72% accuracy, outperforming models based only on demographics. The researchers believe personality profiling using passive phone data could enable large-scale studies but note more validation work is needed.