Research is by definition an attempt to systematically investigate an issue so as to add to our understanding. It is a structured expression of our curiosity. Three papers in this issue sit at the limits of our search for knowledge—what we should know, what we can know and—maybe—how to know a little bit more.

Colleges, understandably, are curious about the students they plan to enrol. I wasn't aware, however, that over 50% of colleges in the USA specifically ask prospective students about their criminal background.1 As Runyan and colleagues have noted, this is driven in part by the belief that a substantial proportion of crime on college campuses is perpetrated by students themselves. If one could predict which students are likely to be perpetrators, this could influence admission decisions or trigger some form of oversight, supervision or remediation. The authors report a strong positive association between precollege criminal activity and college misconduct.2 However, and to their great credit, the authors look beyond the epidemiologic association …