Skip to content

Epistemology 2: The Amazing Success of Statistical Prediction Rules


For the past half century or so, psychologists and statisticians have shown that people who have great experience and training at making certain sorts of predictions are often less reliable than (often very simple) Statistical Prediction Rules (SPRs)(25).

This seems to be a central point of this book. Their epistemology involves these SPRs a whole lot. Science of knowing? I’m still trying to garnish the full depth of what they are saying from the first chapter, but my basis in Standard Analytic Epistemology is not very developed. I had never heard of it before. Let’s move on.

We have coined the expression ‘Ameliorative Psychology’ to refer to the various empirical work that concerns itself with passing normative judgment on reasoning strategies and prescribing new and better ways to reason (26).

Good. A definition.

There is an intuitively plausible explanation for the success of proper linear models. . . they are constructed so as to best fit a large set of data. But typical human predictors do not have all the correlational data. . . as a result, we should not find it surprising that the proper linear models are more accurate than even expert humans. While this explanation is intuitively satisfying, it is mistaken (27).

Whoa, I thought they were giving the explanation before I read that last line. This is getting interesting.

When gatekeepers make judgments about candidates on the basis of a dossier and an unstructured interview, their judgments come out worse than judgments based simply on the dossier (32).

I think this may be where the answer comes about. It’s not that humans don’t have the correlation data available, it’s that they look to the salient details when they make decisions, not necessarily to the pertinent details. An interview is distracting, and is less predictive than previous job performance. But, it is going to be more salient, so a good first impression can trump the evidence.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: