Many experiments in cognitive psychology and other domains ask for confidence ratings (e.g., on a 0-100 scale, 100 meaning "I'm sure I experienced this stimulus"). What accounts describe how these decisions are made?

Review papers of this issue would be particularly helpful.

  • $\begingroup$ Maybe look at the question about forced decisions, and my answer about evidence accumulation: cogsci.stackexchange.com/q/13764/11318 . I guess, but do not know, that confidence is an estimate of how quickly you accumulated evidence for one of the decisions, and how much evidence there is for the other. The process is subconscious though, so how one would make the result conscious and verbalize this, I don't know. Perhaps an actual estimate of how quickly you knew. $\endgroup$ – Robin Kramer Apr 22 '16 at 19:15
  • $\begingroup$ Have a look at this recent Neuron article, it's not a review but might give you a good -though not simple- starting point cell.com/neuron/abstract/S0896-6273(15)00828-4 $\endgroup$ – elisa Apr 24 '16 at 14:53

Note: There are different kinds of confidence ratings (eg, confidence in skill or ability, confidence in knowledge or expertise, etc). This answer only addresses the question as asked, regarding confidence in memory or familiarity.

Processing fluency - sometimes referred to as "cognitive fluency", "cognitive ease", or just "fluency" - is a mechanism proposed to underlie confidence ratings about familiarity:

... perceptual fluency can contribute to the experience of familiarity when fluent processing is attributed to the past. ... Jacoby and Dallas in 1981 argued that items from past experience are processed more fluently. ... people sometimes take fluency as an indication that a stimulus is familiar even though the sense of familiarity is false. Perceptual fluency literature has been dominated with research that posits that fluency leads to familiarity.

Processing fluency, as investigated by Hertwig et al (2008), refers to latency - ie, the speed with which processing occurs. They demonstrate that the human mind is able to discriminate retrieval (response) latencies with great accuracy (down to 100ms), and these latencies correlate strongly with recognition confidence ratings.

Related video.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.