The trouble with studying behavior is that people are much less reliable than mice or bacteria. While you can look at molecules in a Petri dish and measure change, people, on the other hand, are notoriously dishonest.
Two recent experiments make the point. When researchers attached motion sensors to the wrists of sleepers, they found that people may be getting less sleep than they said they did when asked to keep a diary. The researchers – none in rocket science – also discovered that rich people sleep better than poor people. Imagine that!
In another study, researchers tried something new, other than asking elderly people about their exercise regimens. They actually measured output of carbon dioxide, a marker for activity, and found evidence that exercise among the elderly is even more important than earlier thought. Long story short: Over six years, 12 percent of the most active group died, 18 percent of the middle group died and 25 percent of the least active died.
It may be time for social science to do what the television ratings people are doing – measuring with “people meters” not what people say they do but what they actually do.
I’m surprised this conclusion hasn’t become common knowledge. Mice and other animals can be trusted because we study them directly and don’t rely on what they tell us. This not only gives us sound data, but easy replicability. Humans should be no exception.
Actually such studies are not so unusual. For example Pascal Boyer in Religion Explained describes such research on what people actually believe about religious matters in contrast to what they say they believe.