Are you what you eat?
Thursday, April 1, 2010 at 2:15PM 
The BBC (and others) have reported that “TV chef Jamie Oliver's campaign for healthy school dinners boosted pupils’ test results”.
We should always be cautious of confusing correlation (food has become healthier, results have got better) with causation (healthier food causes better results). Just because two things both increase doesn't mean that one causes the other. (And which way round is it? Maybe better results led to the pupils eating healthier food.)
We should also be cautious of what claims actually mean. The BBC report says the results “showed the improvements were down to the effects of the campaign”. Note that this is a slightly different to saying that the improvements in results were down to improvements in the food. It is the campaign which led to the improvements.
So the improvement may be down to the so-called Hawthorne effect, which says that people may improve their performance merely because they know they are part of a project which is intended to improve their performance. This is similar to the placebo effect.
Matthew Handy
Recent studies have called the Hawthorne effect into question. See, for example, this report from The Economist.
hawthorne effect,
placebo effect,
statistics 
Reader Comments