Confidence, uncertainty, and the use of information.

Abstract

When a person makes predictions about an unknown quantity, a distinction can be made between beliefs about possible values for the quantity and the belief that a given prediction is correct. We use the term uncertainty to refer to the former, and confidence to refer to the latter. Four experiments demonstrated that confidence and uncertainty are affected in different ways by the available information. Confidence increased as the amount of information increased, especially if this variable was manipulated within subjects. Confidence was reduced by increasing the apparent difficulty of the task if manipulated within subjects, but not between subjects. Uncertainty increased with the amount of information (i.e., certainty was reduced), a result that is inconsistent with statistical theory. We proposed that uncertainty is determined by the number of different predictions that can be generated, whereas confidence is influenced by salient factors that people believe affect the accuracy of their predictions. Information is used to suggest possible outcomes in the former case, and to evaluate hypotheses in the latter case.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 279 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.