A nearly universal problem in research with human subjects is the impossibility, from both a practical and an ethical standpoint, of keeping research subjects unaware that they are participating in a research project. Unfortunately, the mere recognition that they are being observed can exert a profound inﬂuence on the behavior of research participants. Whereas the usual goal of observational research is to describe behavior as it occurs naturally, it is actually quite difﬁcult to get people to “act naturally” when they know they are being observed. Also, recognition of various features of an experimental design, known as demand characteristics, can cause research participants to form hypotheses regarding what the experimenter is hoping to observe and to alter their behavior accordingly.
This phenomenon is widely known as the Hawthorn effect, named for an industrial-engineering study conducted in 1926 at Western Electric’s Hawthorn works, a telephone assembly factory in the Chicago suburb of Cicero, Illinois. The researchers were investigating the effect of different levels of ambient light on worker productivity. The experiment was performed by reducing the amount of light in the factory over a period of several days. The unexpected result was that the women working in the factory continued to work at the same high level of productivity even when lighting was reduced to the point where they were assembling telephones in near-total darkness. Rather than conclude erroneously that lighting levels were completely irrelevant for productivity, the researchers realized that the women, knowing they were being closely watched but unaware of the research hypotheses, formed their own ideas about what was going on and worked extra hard to make a good impression, rather than acting naturally.
Whereas the Hawthorn effect applies primarily to the experimental group in a study, subjects for whom a condition is changed to observe the effects of the change, some observers have also identiﬁed a related effect that applies to the control group: the John Henry effect. Named for the folk hero who worked extra hard to show his superiority to the new steam-powered hammer, the John Henry effect is the tendency for persons in a control group to attempt to outperform the members of the experimental group to whom a treatment has been applied. This would of course invalidate the purpose of including the control group, and the need to avoid this effect is a major reason for the popularity of single and double-blind experiments.
The Hawthorn study has achieved near-mythic status among both social scientists and industrial productivity consultants, to a point at which tracking down the true story is a bit challenging. Books and Web sites variously attribute the study to both General Electric and Western Electric, and the location is variously cited as Ohio, New York, Pennsylvania, and only occasionally Illinois. The Hawthorn effect is now frequently cited by so-called experts in business productivity as an example, not of research bias, but rather of the role of a caring management style in improving productivity. The explanation offered is usually that the study showed that when employees feel that management cares enough about them to watch them closely, they reward the positive attention by working harder.
- Lehner, Philip N. Handbook of Ethological Methods. New York: Cambridge University Press, 1998.