Processing Linguistic Probabilities: General Principles and Empirical Evidence

Processing Linguistic Probabilities: General Principles and Empirical Evidence

David V. Budescu and Thomas S. Wallsten. 1995. (View Paper → )

How do people use and understand linguistic expressions of probability? Is information processing, choice behaviour, or decision quality more optimal in any well-defined sense when subjective uncertainty is expressed linguistically or numerically? Do people communicate with each other better in one modality or another? These and related questions are of considerable practical and theoretical importance. Practical issues arise because weighty decisions often depend on forecasts and opinions communicated from one person or set of individuals to another.

The english language has plenty of words, but when stakes are high be careful when communicating probabilities that way.

• Use numerical probabilities instead of words like "likely" or "possible" for critical decisions since people interpret these terms vastly differently
• Provide both verbal context and specific numbers when presenting to stakeholders since people prefer giving information verbally but receiving it numerically
• Explicitly define what probability terms mean in each specific context since the same word means different things in different situations
• Treat all estimates as ranges rather than precise values since even "exact" numbers from stakeholders contain inherent uncertainty
• Present high-priority items with concrete data rather than vague language since people weight more precise information more heavily in decisions
• Create standardised ways for teams to express uncertainty and confidence to avoid systematic misinterpretation across functions
• Build buffer time and resources into plans based on the vagueness of human probability communication

What do these mean to you? (Almost Certain, Probable, Likely, Good Chance, Possible, Tossup, Unlikely, Improbable, Doubtful, Almost Impossible).