Effects of contextual cues in recall and recognition memory: the misinformation effect reconsidered

Research in semantic word list-learning paradigms suggests that presentation modality during encoding may influence word recognition at test. Given these findings, it is argued that some previous misinformation effect research might contain methodologies which are problematic. Misleading information groups typically receive erroneous information in written narratives, which may be further impeded by written tests. Results may therefore be explained by misinformation or encoding specificity. In two experiments, participants received restated, neutral, and misleading post-event information through auditory or written modalities. Participants' recognition and recall of critical details about the source event were tested. In a recognition test using the standard testing procedure, there were no condition differences for post-event information presented via an auditory modality. However, for post-event information presented in the text modality, recognition performance was more accurate for restated information relative to neutral information, which in-turn was better than the misled condition. Using the modified testing procedure, the differences were again limited to the text condition. Better performance was evident in the restated condition relative to the average of the neutral and the misled conditions, and there was no difference in performance between the neutral and the misled conditions. Using a recall test, however, there was no effect of modality. Memory was significantly better for restated information than for the average of the neutral and the misled conditions and poorer in the misled condition relative to the neutral condition. Results are discussed in terms of the effects of contextual cues at test, and methodological and interpretational limitations associated with previous research.