Recall Bias

Some observational epidemiology studies, particularly retrospective case-control studies that are designed to assess chemical safety require interviewing subjects about their personal behavior, sometimes expecting them to recall chemical exposures dating back decades. Many times, individuals suffering from a health ailment might exaggerate their exposure level since its natural to look for an explanation for illnesses whose causes we really do not understand. Other times, individuals simply cannot recall well enough for the data to be particularly useful, but researchers may lead them to provide it anyway.(1)  For example, Geoffrey Kabat of the Statistical Assessment Service (STATS) points out how recall bias rendered the results of a number of retrospective studies linking brain cancer and cell phones completely wrongheaded when compared to prospective cohort studies of cell phone users.

According to the literature, recall bias is most significant when a disease under investigation is serious (such as with cancer), the subject believes the risk factor is high, news reports have exaggerated risks about the substance being studied, or the chemical in question is not socially acceptable (such as with illegal drug use).(2) Recall bias can so undermine the validity of the data the final study results are completely off the mark.

Researchers face a number of other challenges to establishing the validity of their findings. These include challenges associated with finding a truly random sample, sometimes insufficient sample size, confounding factors, recall bias, and, researcher bias.

Browse the terms on the sidebar of this webpage for more information and/or download a copy of A Consumer’s Guide to Chemical Risk:  Deciphering the “Science” Behind Chemical Scares.


(1) Geoffrey Kabat, “Yet Another Large Study Discredits The Alleged Link Between Cellphones and Brain Cancer,” Forbes (online), September 4, 2013, r.
(2) Eman Hassan, “Recall Bias Can Be a Threat to Retrospective and Prospective Research Design,” The Internet Journal of Epidemiology 3, no. 2 (2006).

Copy Protected by Chetan's WP-Copyprotect.