Thursday, December 18, 2014

How Academia's Liberal Bias Is Killing Social Science



By Pascal-Emmanuel Gobry
Wednesday, December 14, 2014

I have had the following experience more than once: I am speaking with a professional academic who is a liberal. The subject of the underrepresentation of conservatives in academia comes up. My interlocutor admits that this is indeed a reality, but says the reason why conservatives are underrepresented in academia is because they don't want to be there, or they're just not smart enough to cut it. I say: "That's interesting. For which other underrepresented groups do you think that's true?" An uncomfortable silence follows.

I point this out not to score culture-war points, but because it's actually a serious problem. Social sciences and humanities cannot be completely divorced from the philosophy of those who practice it. And groupthink causes some questions not to be asked, and some answers not to be overly scrutinized. It is making our science worse. Anyone who cares about the advancement of knowledge and science should care about this problem.

That's why I was very gratified to read this very enlightening draft paper written by a number of social psychologists on precisely this topic, attacking the lack of political diversity in their profession and calling for reform. For those who have the time and care about academia, the whole thing truly makes for enlightening reading. The main author of the paper is Jonathan Haidt, well known for his Moral Foundations Theory (and a self-described liberal, if you care to know).

Although the paper focuses on the field of social psychology, its introduction as well as its overall logic make many of its points applicable to disciplines beyond social psychology.

The authors first note the well-known problems of groupthink in any collection of people engaged in a quest for the truth: uncomfortable questions get suppressed, confirmation bias runs amok, and so on.

But it is when the authors move to specific examples that the paper is most enlightening.

They start by debunking published (and often well-publicized) social psychology findings that seem to suggest moral or intellectual superiority on the part of liberals over conservatives, which smartly serves to debunk both the notion that social psychology is bereft of conservatives because they're not smart enough to cut it, and that groupthink doesn't produce shoddy science. For example, a study that sought to show that conservatives reach their beliefs only through denying reality achieved that result by describing ideological liberal beliefs as "reality," surveying people on whether they agreed with them, and then concluding that those who disagree with them are in denial of reality — and lo, people in that group are much more likely to be conservative! This has nothing to do with science, and yet in a field with such groupthink, it can get published in peer-reviewed journals and passed off as "science," complete with a Vox stenographic exercise at the end of the rainbow. A field where this is possible is in dire straits indeed.

The study also goes over many data points that suggest discrimination against conservatives in social psychology. For example, at academic conferences, the number of self-reported conservatives by a show of hands is even lower than the already low numbers in online surveys, suggesting that conservative social psychologists are afraid of identifying as such in front of their colleagues. The authors say they have all heard groups of social psychologists make jokes at the expense of conservatives — not just at bars, but from the pulpits of academic conferences. (This probably counts as micro-aggression.)

The authors also drop this bombshell: In one survey they conducted of academic social psychologists, "82 percent admitted that they would be at least a little bit prejudiced against a conservative [job] candidate." Eighty-two percent! It's often said discrimination works through unconscious bias, but here 82 percent even have conscious bias.

The authors also submitted different test studies to different peer-review boards. The methodology was identical, and the variable was that the purported findings either went for, or against, the liberal worldview (for example, one found evidence of discrimination against minority groups, and another found evidence of "reverse discrimination" against straight white males). Despite equal methodological strengths, the studies that went against the liberal worldview were criticized and rejected, and those that went with it were not.

I hope this paper starts a conversation. Again, this is not about culture-war squabbling — it is about something much more important: the search for knowledge.

No comments: