Long-term feedback and differential feedback effects in nonmetric multiple-cue probability learning

Abstract

The purpose of this research was to study the long-term effects of different types of feedback upon learning and performance of individual human organisms and the sensitivity of such subjects to changes in the validity of information in probabilistic judgment tasks. The subject's task consisted of predicting a binary event on the basis of two or three binary cues whose validity changed on each of five successive days. Subjects received one of three different types of informational feedback about their performance - outcome feedback, cue-event validity feedback, or percentage correct feedback. The results indicated, at least for tasks involving additive cues, that adaptation to new cue validities became more rapid on successive days. The different types of feedback tended to produce different response strategies. However, the different strategies had virtually no effect upon overall judgmental accuracy. Further, there was an apparent interaction between task characteristics and feedback type. This result was found to be consistent with the results of other studies utilizing differential feedback in similar tasks.