Skip to main content

Research Repository

Advanced Search

The effects of absolute risks, relative risks, frequencies, and probabilities on decision quality

Covey, J.

Authors



Abstract

It is important to understand how the quality of people's decision making may be affected by the format used to present treatment benefits. Two experiments compared the accuracy of presenting the benefits of cancer screening tests or vaccines using either absolute or relative risk formats that included baseline risk information. Moreover, the absolute and/or baseline risks were presented using either natural frequencies or probabilities. In both experiments, accuracy was measured by the sensitivity of choices to differences in absolute rather than relative risks. Experiment 1 showed no significant differences in sensitivity between the relative and absolute risk formats when the risks were presented as natural frequencies. Sensitivity was, however, poor in both probability versions. Experiment 2 tested the natural frequency versions more stringently by presenting choices with different levels of difficulty. The author found that decision quality was significantly less affected by increases in difficulty in the absolute risk format. Presenting baseline risks using natural frequencies may help to reduce the biasing effects of relative risks but decision quality may not be on a par with the accuracy of decisions made when absolute risks are presented in natural frequency formats.

Citation

Covey, J. (2011). The effects of absolute risks, relative risks, frequencies, and probabilities on decision quality. Journal of Health Communication, 16(7), 788-801. https://doi.org/10.1080/10810730.2011.561916

Journal Article Type Article
Publication Date Jan 1, 2011
Deposit Date Aug 24, 2011
Journal Journal of Health Communication
Print ISSN 1081-0730
Electronic ISSN 1087-0415
Publisher Taylor and Francis Group
Peer Reviewed Peer Reviewed
Volume 16
Issue 7
Pages 788-801
DOI https://doi.org/10.1080/10810730.2011.561916
Public URL https://durham-repository.worktribe.com/output/1504897