<p>But the problem with those numbers is that MIT is a small school, and although there was a rash of suicides in the late 90s, there weren’t enough, given MIT’s small size, to say that MIT’s average is actually above the national average.</p>
<p>This is especially true because MIT’s student population at that time was strongly male (males are more likely to commit suicide than females) and strongly engineering or business majors (engineering and business majors are more likely to commit suicide than students in other majors).</p>
<p>It is dishonest to wave numbers around without the proper statistical analysis to back them up – sure, the national rate is 7/100,000, but what’s the standard error of the mean for that? MIT’s student population (and the student population at MIT that committed suicide in the late 90s) is too small to draw any conclusions – one statistical treatment I’ve seen concluded that it will not be possible to know whether or not MIT’s average is above the national average for another thirty years. Just because 10 (or whatever) > 7 doesn’t mean MIT’s rate > national average.</p>
<p>Don’t tell me “there’s not a suicide problem because you haven’t seen it”. I’m not interested in anecdotal finger-pointing; I’m interested in a rigorous statistical treatment of the data.</p>