"Was there really a Hawthorne Effect in the original Hawthorne studies?"

This post's title is the question posted today by Steven Levitt on the NYTimes' Freakonomics blog. Ah, the Hawthorne effect, whose mere mention often sends shudders down the spine of veteran AP Psych teachers who recall this term appearing out of the blue on some mid-'90s exam. Since then some (including your correspondent) often feel the need to almost shout it out the door as the students are leaving to take the test -- "and don't forget the Hawthorne Effect was ...!"

Levitt and John List took a new look at the data which originally suggested that any change made at the Hawthorne plant, particularly in terms of lighting, led to an increase in productivity. The conventional wisdom gleaned from the study was that scientists need to be cautious about unintentionally interfering with an experiment because the mere fact that subjects know their behavior is being observed would make change their behavior, and in this case, work harder. The economists re-analyzed the old findings and found that the changes themselves did not make productivity increase but that the differences emerged because the changes in production were compared to the wrong standards. Here are two examples:

• Changes typically took place on a Sunday and then the productivity of the workers was compared from the day before (Saturday) to the day after (Monday). But when the data from Mondays alone was dissected, in turns out that every Monday had greater output that every Saturday -- workers tended to be most productive early in the week and then slack off as the week progressed. (No wonder since they had a six day work week!)

• Another observation found that worker productivity went down significantly after the experiments ended, leading to the conclusion that the presence of the researchers increased output and their absences led to a decline. But what Levitt and List found was that the research stopped in the summer, and every year there was a decline in production during the summer, perhaps because of the heat.

You can read more about this work in this brief Economist article. The full paper is available here but it looks like it'll cost you $5. A nice online exhibit from Harvard Business School is here.