Related Experiment Videos
Characterizing physicians' perceptions of ambiguity
S P Curley1, M J Young, J F Yates
1Department of Management Sciences, University of Minnesota, Minneapolis 55455.
Abstract:
Ambiguity, typically characterized as imprecision in judged probabilities is distinguishable from that uncertainty typically measured using probabilities. It reflects the inadequacy of a point probability judgement and the assumptions upon which it is based. Ambiguity influences both patients' and physicians' decisions, but it is unknown how decision makers conceptualize ambiguity. Toward this goal, properties of three ambiguity indicators were examined: a confidence rating, a plausible range, and an interquartile range. Board-certified internists and fourth-year medical students evaluated simulated cases of suspected coronary artery disease. Their judgements provided insights into the three ambiguity indicators. The distribution-based interquartile range was largely redundant with the plausible range, and was least adequate. The confidence rating was not equivalent to the plausible range, and appeared to best reflect the construct of ambiguity as it has been defined.
Insights
Physicians
Area of Science:
- Decision analysis in medicine
- Medical decision-making
- Cognitive science in healthcare
Background:
- Ambiguity, distinct from uncertainty, impacts patient and physician decisions.
- Understanding how decision-makers perceive ambiguity is crucial.
- Existing measures of ambiguity in medical judgment require further examination.
Purpose of the Study:
- To investigate how ambiguity is conceptualized by medical professionals.
- To evaluate the properties of three distinct ambiguity indicators: confidence rating, plausible range, and interquartile range.
- To determine which indicator best reflects the construct of ambiguity in clinical judgment.
Main Methods:
- Utilized simulated cases of suspected coronary artery disease for evaluation.
- Involved board-certified internists and fourth-year medical students as participants.
- Assessed the properties and adequacy of confidence ratings, plausible ranges, and interquartile ranges as measures of ambiguity.
Main Results:
- The interquartile range was found to be largely redundant with the plausible range and least adequate.
- Confidence ratings were not equivalent to plausible ranges.
- Confidence ratings appeared to be the most effective indicator of ambiguity as defined in the study.
Conclusions:
- Confidence ratings emerge as a superior measure for reflecting ambiguity in medical decision-making.
- The plausible range is a more adequate indicator than the interquartile range.
- Further research into ambiguity measurement can refine clinical decision support tools.