An anonymous IWF author, criticizing Mary Koss’ famous study of rape incidence (there’s more info in this post, for those of you who have no idea what I’m talking about), wrote:
Although the IWF doesn’t say what the purpose of presenting this statistic is, most readers will take it as an indication of how outrageously high Koss’ numbers are. (Indeed, if that isn’t the point, why on earth include the statistic in a critique of Koss at all?)
Such a comparison is wildly unscientific and irresponsible. No legitimate comparisons can be made between a lifetime prevalence self-report survey and the numbers of such incidents that occur on campus in one year. But for the purpose of rebutting the IWF’s bad methodology, I’m going to pretend they can be compared. And unlike the IWF, I’m going to do the math. How far apart are this stat and the numbers given in Koss’ report?
To start with, we need to ask “how many female undergraduates are on an average college campus?” (Koss’ statistic about rape victims included only female undergrads). According to The Chronicle of Higher Education, in 2000 there were 12,450,587 undergrads in the USA. Of those, 55.9%, or 6,959,878, were women. There are 4,096 colleges and universities in the US, which works out to an average of 1,699 female undergrads per college campus.
So of those 1,699 women on a college campus, how many will be victims of rape in a typical year, according to Koss’ numbers? The famous “1 in 4” figure is a lifetime figure that includes both rape and attempted rape; the “less than one per campus” figure, on the other hand, is about completed rapes within a single year. So we can’t compare those two numbers directly.
Fortunately, Koss’ report includes a calculation of annual incidence of completed rapes only – out of 3,187 undergraduate women Koss surveyed, 207, or 6%, were victims of rape in the past year. Since the average US campus has 1,699 female undergrads, and 6% of 1,699 is 102, according to Koss’ study we’d expect 102 undergraduate women at an average college to be victims of rape each year.
Look again at the statistic the anonymous IWF author is comparing Koss to: “…campus police reported 1,310 forcible sex offenses on U.S. campuses in one year.” So we’re not talking about all the rapes that happen in a single year; just the rapes that happen “on U.S. campuses.”
Koss’ survey asked rape victims if the rape had taken place on or off campus. 14% reported that it had taken place on campus. So out of 102 rapes happening to undergraduate women in a year, according to Koss’ results, we’d expect 15 to actually take place on campus.
But we can’t assume that all 15 reported the incident to campus police. According to Koss, only 5% of rape victims said they had reported the incident to the cops at all. And 5% of 15 is 0.75.
So according to the statistic the IWF offers – a statistic that, clearly, is meant to discredit Koss’ results – there is “an average of fewer than one rape per campus” per year that campus police know about. And according to my calculations, based on Koss’ numbers we’d expect to see – less than one rape known to the cops per college campus per year.
In other words, Koss’ study found exactly what it should have, according to the statistic the IWF author supplied. So how, exactly, does that disprove Koss?.
So the IWF has actually proven that they are not only morally suspect, but that they are incapable of reading descriptors and are unable to understand statistics or even arithmetic. But they are darn good at misrepresentation.
In other words, Koss’ study found exactly what it should have, according to the statistic the IWF author supplied. So how, exactly, does that disprove Koss?
If you hold your breath until you see spots and then inhale gasoline fumes, it will all start to make sense to you.
Amp just illustrated that this is a case of what I generally assume:
When someone presents statistics to compare “apples to oranges” , to use the idion, I assume one of three possible things:
1) They don’t know that the difference between apples to oranges. (They are ignorant in one way.)
2) They don’t know that comparing apples to oranges doesn’t make sense. (They are ignorant in a different way.)
3) They know there is a difference, and they know they should compare, but they wish to use non-sense to support their arguments.
Generally, in academia, the correct response is to ignore the ridiculous claim. Unfortunately, once these things get on the news, someone needs to show the comparison is an apple to oranges comparison, explain why the comparison makes no sense and provide the actual “apple to apple” comparison.
Quiet Excellently Done, Amp
This is one of the reasons I read Alas. Excellent, Amp!
Thanks, Amp. I wish that there wasn’t so much disinformation out there that you need to take time out of cartooning to clear it up. If people could just stick to the truth and not distortions, we’d see more cartoons!
You rock.
Verry nice, amp! I so wish that there was some way of requiring peer reviewing of all these think tanks. They are riding on the general idea that research is reviewed, and they are destroying the reputations of other researchers by publishing rubbish.
good number work. “known by campus police” and “reported” may not be the same number.
there’ve been a series of exposes about pressure to under-report.
when my ex was raped off campus in 1981 by muhammad zabeeb, a fellow CU student, she told the city police, who refused to take a report or do anything; we don’t know why.