When a bunch of things are correlated, one of which is time

October 13, 2011 § Leave a comment

We know this is a fallacy:

  1. X has increased over the last 50 years.
  2. Y has increased over the last 50 years.
  3. Therefore either X causes Y, or Y causes X.

If we add an exit clause like “or there’s confounding” to 3., we weaken the argument to uselessness.

Now, although we can’t eliminate the possibility of confounding, we can get interesting evidence if there’s more to the data than “both increase”. If the peaks and troughs are simultaneous, then there’s some kind of strong relationship between the variables, whether causal or not. If one variable consistently leads the other, this can suggest direction of causation, though it’s easy to kid ourselves.

Barring such clear pattern, we need a more complicated causal model. It doesn’t have to be too complicated: something as simple as “X causes Z, which in turn causes Y” is interesting enough to have implications, if true, and is somewhat more susceptible to falsification. We can add more detail to the model as necessary. But you need to be explicit about pathways. Draw a picture if necessary.


Why We Fight

December 26, 2010 § Leave a comment

From the FAQ on the website of the Equality Trust, an organisation founded by Billy Kerry, Richard Wilkinson, and Kate Pickett. The latter two are authors of the highly influential book The Spirit Level.

Correlation does not prove causality. What reasons are there for thinking that inequality is the cause of such a wide range of health and social problems?

To suggest that these relationships are causal does not involve a major departure from what we know already. Within countries we know that all the components of our Index of Health and Social Problems are strongly related to social status: the further down the social ladder the more common they become. The new part of the picture is simply that if you stretch out the social status differences all the problems related to social status become more common. Rather than postulating entirely new causal processes we are therefore only providing a bit more information about the relationships that have always been recognised.

People who have studied the graphs on this web site and in The Spirit Level carefully will have noticed that there is a clear tendency for countries which do badly on one outcome to do badly on others. We show evidence that 10 or 12 different problems tend to move together. That implies that they share an underlying cause. The association between inequality and our Index of Health and Social Problems is very close and no one has yet suggested an alternative.

Lastly, as the different chapters in our book show, many of the causal processes leading from inequality to the various health and social problems are already known. For example, the effects of social status on health have been demonstrated among monkeys in experiments which kept diet and material conditions the same while altering social status by moving animals into new groups and the effects of chronic stress on the immune and cardiovascular systems are increasingly well understood. Similarly, violence is more common in more unequal societies (where status competition is intensified) because it is so often triggered by people feeling looked down on, disrespected and humiliated.

Where Am I?

You are currently browsing the correlation does not prove causation even if you really really want it to category at "But it's under .05!".