Saturday, January 21, 2012

Variance

If someone fills out his All-Star Ballot with 8 Dominican players year after year, it is fair to conclude that said origins explains all the variance in his selection.  We also need look no further for explanations.  Whatever he says about fielding percentages, clubhouse leadership, or whatever is mere pretense.  He votes for Dominicans.  In seeking to explain a behavior (or any phenomenon, actually) we look for what explains the variance.

Caution #1: sometimes there are threshold criteria which are not immediately visible.  In the example above, the voter is choosing among players who have already made it into the major leagues, a pretty select group.  He's not voting for just any Dominican he might meet on the street.

Caution #2: there are a lot of really good Dominican ballplayers.  It could happen that some year they did have a legitimate contender for best at every position.  That's why I added in the "year after year" part. I suppose if one wants to be technical, we should be alert for such possibilities in all our other explanations of variance as well.  Though that seems tiring.

Excluding such hidden factors and statistical rarities, we are wise to simply disregard anything a person says by way of explanation.  It might trick us into believing he is actually thinking.  Some folks are quite persuasive, and we might come away saying "I don't buy everything that he says, but he made some really good points about the Keystone Pipeline."  He made no good points about the pipeline that you couldn't get in much better form from some more thoughtful source. (Note that I have not expressed an opinion on that.  I did it on purpose, to illustrate the point.  The person made no good points about the pipeline. He may have repeated something worth knowing, but you have no way of telling.)

My wife keeps up with local politics much more than I do, contacting representatives and asking them what they think.  One person who has been elected from here for a million years, when confronted with an issue he clearly hadn't thought much about, answered that he supported whatever the Republican Party was saying.  No, he didn't say it so baldly, but that was the essence.  General Republican sentiment explains much of the variance in his opinions.  He may have some opinions where he goes against that grain, but his default is to generic GOP.  Not a person you should ask why such-and-such is a good policy. He would give a good answer only by accident.

At the other extreme, what about the voter that checked off no Dominicans?  We would have a suspicion of anti-DR prejudice, and go looking for explanations of that. Does he resent them and favor Venezuelans instead? Or whites-only?  (Good luck with that team.) For our purposes here, it doesn't matter.  This is another unreliable person.  Don't ask them why they think something.  You already know why.

But if our All-Star voter checked off five Dominicans, we might think that was a little extreme, that he might be a little bit biased, but he is probably worth listening to.  Five's about right. We might be tempted to automatically think he is a reasonable person.  No, we have to check which Dominicans he voted for and why.  He has not yet earned a place at the table, but only at the door.

2 comments:

Anonymous said...

seems to me that if i examine 100 republican/democratic policy differences and find that i agree with the republicans on 95 of them, that i could reasonably default to the republican position on subjects that i hadn't had time to study.
might change my position later, but that's another issue.

Cheap Diablo 3 gold said...

Once more wonderful submit. You seem to have a good understanding of these kind of themes or templates.After i entering your site,I thought this kind of . Seriously and blogging your site is often more interesting. In your Achievement!