Wednesday, July 15, 2015

When no one cares

Sociologists don't get nearly as much attention in the media as other social scientists (Justin Wolfers gives some data and Syed Ali has a more general discussion).  Many eminent sociologists have tried to explain this situation.  On the surface, they offer a wide variety of answers, but at a deeper level they agree:  it's because too many sociologists don't do the sort of research that the author does.  If the author does qualitative work, the reason is that there's too much quantitative research; if the author  does quantitative work, the reason is that there's too little quantitative research (a variant is that sociologists have poor quantitative skills, so when we do quantitative research it isn't very good).  If the author favors "engaged" scholarship, the reason is that too many sociologists try to be above the fray and avoid talking about things that matter to people; if the author favors "scientific" scholarship, the reason is too many sociologists offer nothing more than political advocacy.

I'm going to take another approach and ask what kind of social scientific research gets attention in the media.  I haven't made a systematic count, but it seems that journalists are very fond of two related types of studies--those that find that an apparently small or irrelevant factor has a big influence on something important (e. g., unconscious gender stereotypes influence hurricane deaths!) and those that find a small intervention has a big impact (e. g., a 15 minute conversation leads to lasting change in support for same-marriage! and even affects the opinions of other people in the household!).  You can see why these would appeal to journalists:  there's a story that's easy to tell and surprising.

Sociologists are inclined to think that everything is connected and that any important  social condition has deep roots in social structure and/or culture.  We also tend to be reluctant to put that aside and try "cute little models that one hoped yielded surprising insights" (as Paul Krugman described his own style of work).  As a result, sociologists are less likely to come up with the kind of findings or claims that appeal to journalists.

This inclination has some good consequences.   First, it tends to prevent truly lame-brained research. More important, it encourages us to be serious about considering alternative explanations and controlling for other variables.  A sociologist is not going to ask "does my theory suggest a reason to control for variable z?"; he or she is going to ask "if I don't control for z, will somebody complain that I should have?"  The answer to that second question is probably yes, so the sociologist will control for z, and while they're at it maybe take a look at what happens with the log of z.....

But the inclination also has some negative consequences.  When it comes to generalizations, sociologists have a tendency to go for  "no difference/no change" or "no simple generalization is possible."  These shade into each other:  often it's no definitive evidence plus the principle that  "no difference" or "no change" should get the benefit of the doubt.*  With policy issues, sociologists have a tendency to say that you can't have a real solution  without changing a lot of other things, which would mean a fundamental transformation of society.  As a result, we avoid making gross mistakes, but miss the opportunity to make a contribution.

*This principle can be methodological (an appeal to "Ockham's razor"), or political/social (you shouldn't lull people into complacency by suggesting that things are getting better).

3 comments:

  1. David:

    These are interesting points. I'd like to add another which is the news hook. Political science research tends to be more gradual in focus--in poli sci, we don't do so many "cute little models," actually I can only think of a couple of examples, there's the controversial Christakis and Fowler claim of contagion of obesity, and of course that recent fraudulent paper on persuasion.

    When poli sci gets in the news, it's because _politics_ is in the news. Even then, I've been known to complain that political scientists don't get listened to enough. (The media tends to use M.D.'s as experts on medicine; and economists as experts on economics; but when it comes to politics, they rely on journalists, retired speechwriters, and retired politicians. But that's a complaint for another day.)

    Anyway, I don't have any overarching point, just that a lot of journalism runs on schedule--baseball news in April and October, campaign news as the election is approaching, etc.

    ReplyDelete
  2. Yes, I agree that's another part of the explanation--politics and economic conditions are regular features of news coverage, and political scientists and economists have an inside track on those topics. On models, I meant that in a broad sense of saying "here's a simple idea that explains a lot of things." Sociologists seem to be reluctant to do that. For example, I find it hard to imagine a sociologist writing "Bowling Alone."

    ReplyDelete
  3. This is great stuff, David. Sociologists' reactions to being mostly ignored take the form of being in denial (i.e., thinking people care) or righteous indignation (people would care if they were enlightened)

    Brad Wright

    ReplyDelete