Thursday, April 11, 2013

Stupid, horrible people are not interesting; intelligent, decent people doing stupid, horrible things are.

Some recent activity here at Stat Views has got me thinking about a fairly obvious distinction that still somehow has a way of getting lost

Most bloggers (myself included) spend a lot of time singling out someone for the idiotic or offensive. These posts are fun and the targets often have it coming but there are two things to remember (and God knows this applies to me as much as anyone):

1. It's not that difficult to find an idiot or jerk if you're really looking;

2. There is nothing particularly notable about an idiot acting stupid or a jerk acting obnoxious.

What's interesting and potentially important is when someone who's not at idiot acts like one (analogous arguments hold for jerks and scoundrels for the rest of this post). This isn't just a case of hating the sin and not the sinner; it's more a question of causal reasoning. When a stupid person does something stupid it requires no explanation but when an intelligent person does something stupid (or better yet, engages in a pattern of stupidity), it suggests that something happened to cause a deviation from the expected. That deviation begs a cause.

For example in the 2000 election coverage, the journalistic lapses mentioned previously are interesting because the journalists involved were both professional and highly respected and had risen to the top of a very competitive profession. If a group of high school reporters had propagated errors in a school election and had biased their coverage because of social cliques and because one candidate had given them small gifts and compliments that would not be of any real interest.

What made this interesting and, to use an often misused phrase, significant was the fact that given the initial assumptions we would normally make about these respected and experienced journalists, what we saw was extremely unlikely. So unlikely it demanded an explanation

In this case various social psych phenomena did fit what we observed. Authority figures within the journalistic click were strongly opposed to Gore. Reciprocation meant that Bush's special treatment of the press corps would be returned. Bush's likability relative to Gore meant that he was likely to receive preferential treatment. Social norming meant that these behaviors would be internalized and repeated

Just to spell things out, what's interesting here is the way that one hypothesis (the Washington press corps was professional and impartial) fit the data badly while another (social dynamics were influencing journalists) fit the data well.

Sometimes what's interesting is neither the person or the act but the reaction to it. For example there was nothing that out of the ordinary about Timothy Noah complaining about the boss who fired him, there was however something strange about the way Politico reacted to Noah's complaint

By the same token there's nothing particularly interesting about a high school student writing a sarcastic essay about the schools that rejected her application. What is interesting is the way that essay illustrates an anticipated reaction among people of a certain social class to changes in the way students are admitted to prestigious schools.

It's easy for the reader (and too often the writer) to lose sight of what's going on. If they rise above the look-at-the-moron humor, posts about stupidity generally need to be a kind of significance test. Start with the assumption that "complete idiots are rare" and seldom make it through demanding selection processes, then, if you are faced with extreme stupidity, ask yourself how likely it is that someone that dim would have reached his or her position. At a p value of around 0.01, it's time to start looking for alternative explanations.

[note: I tried something new with this post, dictating the first draft to my I-phone. I found if I spoke very slowly and enunciated every syllable, the results were acceptable though I did get a couple of odd errors like cherlist for journalist (and, yes, 'cherlist' is a new one on me). I believe I caught all of the Siri-isms but if you see a complete non sequitur, it was probably supposed to be something else.]

No comments:

Post a Comment