<p>This is old news, but still a valid point–my undergrad advisor once remarked that “psychology is actually the study of educated, [mostly] white men [and women these days], ages 18-22.” OTOH, you the process of research is and of itself selecting. We see this in the studies we do on interpersonal violence (in non-college samples)–we get (likely) somewhat inflated rates of abuse experience in those studies, even we don’t directly advertise them as “abuse” or “violence” studies, for participant safety ratings. As another example, when you do studies involving a substantial time commitment, you’re more likely to select in people who are older (retired) and/or low income (have time and need even the “non-coercive” payment for participation). I’m not saying the impact of using a lot of college students should be ignored (it shouldn’t, though I’d note that most of the studies I’ve personally worked on have either been looking at college students specifically or involved community samples. My experience is probably not representative, though), but that self-selection is an issue with any method of recruiting/sampling.</p>