Build Your Own Baloney Detector

A tool-kit for avoiding being fooled

Monday, July 30, 2012

An Interesting Example of a Context-Free Statistic

Here’s an article about an out-of-court settlement between a firm that was collecting payments on behalf of a hospital system and the State of Minnesota. The details aren’t of interest, here, except this: the firm is paying $2.5 million to settle the suit. Got that?

So, I have a vague idea of how much $2.5 million is. You most likely to do, too. But it’s still a low-context (or context-free) statistic. Why? Because we don’t know where the money will go, first of all. Will it go to pay back people who were wrongfully harassed? To pay people who were wrongfully charged money that they then paid? Or is this a punitive payment to punish the firm for their actions? In any of these cases, we’ll need more information to assess how much $2.5 million really is.

Consider: if we’re in the first case, then I’d want to know how many people will get the money and how this average payment compares to other such payment. In the second case, how does the total compare to how much people were wrongly charged? If $2.5 million is much less than that figure, this payment is symbolic at best. If it’s a lot more, then clearly — even with interest — this money is for more than just paying back those wronged.

And in the final case, I’d really want to know what this firm’s profits were like (at least in Minnesota). If $2.5 million is compared to $1 billion in profits, well, that seems like a very different punishment than $2.5 million compared to $10 million in profits, doesn’t it?

I can’t think of a case where I don’t need to know what this money is for and how it compares to some other related figure. We need context!

posted by John Weiss at 18:17  

No Comments »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a comment

You must be logged in to post a comment.

Powered by WordPress