Friday, April 29, 2011

Invalid syllogisms

The Wikipedia entries classify post hoc ergo proctor hoc and cum hoc ergo proctor hoc as logical fallacies, but I suspect this itself is an error.

The categories "true" and "false" belong to statements, while the categories "valid" and "invalid" apply to arguments.

In my recent writings on customer service and attending an event on the same I've been fascinated with the efforts that show a correlation between a measure of customer satisfaction and a measure of profitability. This is especially true when that measure is something like number of products purchased, or lifetime value.

The correlation itself is unsurprising - after all if I had decided to buy more would I say I wasn't satisfied. The question though is one of causation. Is there some magical process whereby I can "manipulate" your satisfaction and this then causes you to buy more?

the answer is clearly no. In reality the two measures are probably correlated because they have common causes.

Novae Meridianae Demetae Dexter delenda est

No comments: