Unacceptability bias

A systematic difference in response rates or uptake of tests due to their “unacceptability”

Background

David Sackett listed unacceptability bias in his 1979 paper documenting sources of bias in research studies, and provided the following explanation: “Measurements which hurt, embarrass or invade privacy may be systematically refused or evaded.”

Example

The collection of information about street drug use could be affected by unacceptability bias, with higher users of street drugs potentially less likely to report their use accurately than lower users, due to the perception that the questions about use are too invasive. We have not found reports of unacceptability bias per se in the literature.  Alcohol under-reporting is one example?  E.g. Drinking pattern is more strongly associated with under-reporting of alcohol consumption than socio-demographic factors: evidence from a mixed-methods study and Underreporting in alcohol surveys: whose drinking is underestimated. And also sexual behaviours presents methodological challenges in survey research due to unacceptability bias.

 

Impact

We have not found any studies evaluating the impact of unacceptability bias. The role of selection bias is discussed in our catalogue.

Preventive steps

Avoidance of harm should be a priority in any research work. Collection of data that is perceived to be personal or sensitive has ethical implications and requires careful planning and administration to be appropriate to the needs of study participants.

Sources

Sackett DL. Bias in analytic research. J Chron Dis 1979; 32: 51-63

Fenton KA et al. Measuring sexual behaviour: methodological challenges in survey research. 

PubMed feed