Recognition of added and deleted details in scripts.

Recognition of script actions that varied in specificity and expectancy of details was investigated. In Experiment 1, subjects indicated whether each action was the same or changed on an immediate or delayed yes/no recognition test. Changes that involved added details were recognized better than changes that involved deleted details. Unexpected added details were detected better than expected ones, but expectancy had no effect on deleted details. Experiment 2 tested whether the poor recognition of changes in actions with deleted details was due to a failure to retrieve those details. The recognition test was a forced-choice test with details present in the correct alternative, so their retrieval was not necessary for correct choices. Still, recognition of originally generic actions was better than recognition of originally detailed actions. Thus, a failure to retrieve details could not completely explain the results of Experiment 1. The subjects probably recognized originally generic actions better because they processed the material schematically so that the detailed actions subsumed the generic idea. Recognition decisions may then have been based either on the plausibility of the alternatives or on their familiarity within the experimental context.