Weighing the Evidence: The Hawthorne Effect

Abstract

The deplorable working conditions in factories across the United States came to national attention in the early 1900s. By the 1920s, corporations were interested in the link between workers' productivity and their working conditions. A series of experiments were conducted from 1924 to 1932 at Western Electric's Hawthorne Works, a large factory outside of Chicago that produced electromagnetic relays for telephone switchboard assemblies. Productivity was measured under various working conditions and job satisfaction assessed by thousands of interviews. The vast data collected remain today the subject of scholarship on organizations, leadership, and motivation. 1,2The first of the studies examined the effects of lighting conditions. Although there was no correlation between the workers' efficiency and lighting level, productivity increased for a short time when the lighting level was changed, irrespective of whether it was brighter or dimmer, but returned to baseline when the study period ended. Similar changes in the workers' behavior were observed when other changes were made -eg, the work area cleaned or relocated. Such transient improvements in behavior in response to being the subject of a study, irrespective of the specific intervention, was termed the Hawthorne effect by Henry A. Landsberger in 1955. 2 The Hawthorne effect has been recognized in medical studies of topics as varied as hand washing and erectile dysfunction. 3,4 The chance that a Hawthorne effect will bias results is highest when healthcare providers or patients understand that they are being observed for a study for a brief period. Readers should consider the possibility of bias when assessing the results of this type of study. The Hawthorne effect can be minimized by conducting observations over a long period and by the use of double-blind trials in which both providers and patients are unaware of either the intervention of interest or the placebo or dummy intervention.

Challenges to Shared Decision Making About Treatment of Advanced CKD: A Qualitative Study of Patients and Clinicians

“…This technique has advantages over direct observation of clinical encounters, which can influence patient and clinician behaviors and be unreliable because it is not guaranteed that treatment of advanced CKD will be discussed during any given observed encounter. 49 However, documentation is also limited by what clinicians chose to document and may not accurately or completely reflect patient-clinician interactions. Third, we cannot be certain that patients and clinicians were referring to the same encounters during interviews or to distant encounters that would be subject to recall bias, both of which might contribute to the observed differences in perspectives between patients and clinicians.…”

Evaluation of a Faculty Development Program Aimed at Increasing Residents' Active Learning in Lectures

“…18,19 Our study has several limitations. They include a singleprogram intervention, limiting generalizability; voluntary workshop participation, which may result in selection bias; faculty behaviors that may be influenced by the presence of observers (a Hawthorne effect) 20 ; and observers who were not blinded to the purpose of the study, introducing the potential for observer bias.…”