February 15, 2002
From the desk of Jane Galt:Benjamin Kepple has a different
Benjamin Kepple has a different point of view on the Horowitz poll:
To prevent any Krugman-esque follies, let me say this: I am a former employee of Mr Horowitz's. I also spent six years, four in the collegiate arena and two in the professional arena, reporting and commenting on the, ah, unique political situation in academia. That experience, however, lets me say without fear that Mr Horowitz's poll is not "discovering" a new phenomenon, it is not suggesting it, it is not hinting that bias may exist in academia. For he knows -- as everyone does in their hearts -- that a fearsome political bias exists in academia.What his poll does is take the reality everyone knows about and tries to put it into the stark, black-and-white realm of statistics.
I agree with Kepple that anyone who thinks that a rigorous poll would deliver significantly different results is sucking too hard on that bong. Nonetheless, I think it's important to try to do a rigorous poll along the lines I outlined for several reasons. The first is that however well you think you know something anectdotally, you don't know it until you can prove it. The second is that a rigorous study would give us the data to begin defining the precise scope and nature of the problem, begin analyzing possible root causes, and examine potential solutions. For example, right now there may be hiring bias, but there isn't any way to show it experimentally. On the other hand, if we could correlate the voting preferences of people extended offers with the voting preferences of the hiring faculty, that would be suggestive, particularly if the group receiving offers were different from the applicant pool. And the third reason is that there is a vast pool of people, in academia and out, who currently deny the scope of the problem. Only massive amounts of data from well designed studies will overcome their objections. I understand that the hard left will question any data, no matter how good, but there will be some towards the center who will be convinced, which is a major step towards rectifying the problem. At least, insofar as it can be rectified.
This is not to criticize Horowitz. His poll was as good as any journalist's poll ever is, and better than some. It is nonetheless fatally flawed because it is a reader response poll, which almost always introduces selection bias. For example, Nader voters may have declined to sent it in because they didn't want to give Horowitz grist; conversely, they may be more ideological and hence more predisposed to send it in. We don't know unless we do a better poll. So I think that Horowitz's poll is suggestive; moreover, I think it is probably approximately right. But I won't know how right it is, or why the phenomenon occurs, or what I can do about it, until I get better data. Horowitz has window on academia; it's up to others to open the door and step through.
Posted by Jane Galt at February 15, 2002 06:26 PM | TrackBack | Technorati inbound links