What is Membership Bias?
Membership bias is when a group of people tend have a specific characteristic that can affect a study’s outcome. Assignment to the group is based on this particular characteristic.
In order for a characteristic to create membership bias, group membership can’t be random in any way (Gabay, 2015). For example, membership to the group “people with lupus” requires the person to have a lupus diagnosis (the “characteristic”), and assignment to the group “people with lupus” is based on that diagnosis. However, if you take that group and randomly assign them to a treatment group or a control group, those individual groups no longer have membership bias.
Effect on Generalizability
When you’re performing a study, choosing a sample from a specific group of people may mean your results won’t be generalizable to the population as a whole. For example, if your longevity study samples from gym membership lists, your results will be biased towards those people who are more healthy in general.
The Healthy Worker Effect is a type of membership bias. Membership in any group of “people who are working” also means the exclusion of less healthy people, including those with chronic illnesses and disabilities. If you study doctors, nurses, or any other group of professionals, you’ll not only be sampling from a healthier group; you’ll be only including those people who are more educated. Bias sneaks in because people with more education tend to live healthier, longer lives than those people with fewer years in school (VCU, 2015).
Membership bias is a term used almost exclusively in epidemiological studies.
Gabay, M. (2015). The Clinical Practice of Drug Information. Jones & Bartlett Publishers.
Gail, M. (2000). Encyclopedia of Epidemiologic Methods. John Wiley & Sons.
Meyer, D. (2004). Essential Evidence-Based Medicine, Volume 1. Cambridge University Press.
Virginia Commonwealth University Center on Society and Health, (2014). Education: It Matters More to Health than Ever Before. Retrieved December 8, 2014 from: https://societyhealth.vcu.edu/work/the-projects/education-it-matters-more-to-health-than-ever-before.html
If you prefer an online interactive environment to learn R and statistics, this free R Tutorial by Datacamp is a great way to get started. If you're are somewhat comfortable with R and are interested in going deeper into Statistics, try this Statistics with R track.Comments? Need to post a correction? Please post on our Facebook page.