Communicating certainty via verbal probability phrases: comparing health contexts with no context

Yiyun Shou*, Lok Him Lee, Joey Elizabeth Yeo, Michael Smithson

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

Introduction: Medical professionals often use verbal phrases to communicate uncertainties and certainties with their patients and the general public. Objectives: This study aimed to investigate factors that can influence people’s interpretation of probability phrases of certainty and uncertainty communicated by doctors in health and medical settings. Methods: An experimental study with a randomized factorial design was conducted to examine both context-related factors and individual difference factors on participants’ interpretation across directions of phrases and frames of the context. Results: Context significantly influenced participants’ interpretation of probability phrases regardless of their level of certainty. Participants’ self-reported prior beliefs were the main driver of this context effect. When participants were in a state of uncertainty, their reliance on experts significantly reduced the prior beliefs’ effect. Finally, refuting probability phrases might provide more reassurance than affirming phrases. Conclusion: People may perceive the outcomes communicated by doctors differently from what the doctors intend to convey, depending on their prior beliefs, reliance on doctors and the direction of phrases. When communicating uncertainties and certainties with patients and general public, it is important for medical professionals to understand audience’s prior experiences, and to foster medical trust and audience’s reliance on medical professionals to reduce bias in clinical risk communication.

Original languageEnglish
Article number14
Number of pages10
JournalBMC Primary Care
Volume26
DOIs
Publication statusPublished - 2025

Fingerprint

Dive into the research topics of 'Communicating certainty via verbal probability phrases: comparing health contexts with no context'. Together they form a unique fingerprint.

Cite this