Additivity neglect in probability estimates: Effects of numeracy and response format

Abstract

When people are asked to estimate the probabilities for an exhaustive set of more than two events, they often produce probabilities that add up to more than 100%. Potential determinants for such additivity neglect are explored in four experiments. Additive responses vary between experimental conditions, mainly as a result of response format, with a scale format leading to fewer additive responses than a list format with self-generated, written probabilities. Participants with high numeracy scores produced more additive responses, especially after being primed with a numeracy scale. Additivity neglect for 100% sums appears to be unrelated to other subadditive judgments, like non-additive disjunctions. We conclude that additivity neglect is caused by a case-based approach, which comes natural in real-life situations where the full set of outcomes is not available.