Last Friday's edition of The Globe and Mail newspaper included a story on how predictive analytics might help decrease university drop-out rates. Several Data Science companies, including IBM’s Watson Analytics and Microsoft, are offering universities their expertise on identifying students at risk of dropping out. Of course, having low marks is a good predictor, but it is not the only one. Women, for example, are less likely than men to leave university before finishing their studies, but when they do, their marks are usually quite good. So why are they dropping out?