| Abstract: | From a cognitive perspective, informative feedback is assumed to be confirmatory or corrective. However, feedback is not always effective and errors persist. In cases where errors persist, an analysis of errors as well as correct responses is recommended in order to better understand the processing of informative feedback. A model of multiple-choice processing is introduced to provide the context for a consideration of the role of error analysis. Data from experiment 1 demonstrate the use of semantic retrieval cues in multiple-choice performance. While informative feedback improves post-test performance, it is difficult to demonstrate a differential effect for type of feedback. Experiment 2 suggests a threshold effect for informative feedback. Information beyond that which is necessary for the confirmation of a correct response and/or the correction of an error has a deleterious effect on post-test performance. Error analysis indicated a disproportionate number of perseverative errors when the use of a retrieval process was not required at post-test. |
