Adult age differences in memory for metaphoric sentences : the role of imagery and semantic feature models
Date
1989
Authors
Fox, Diane Patricia
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
On the basis of theories of metaphor processing, two experiments were conducted on the retention of metaphoric sentences by young and old adults. Experiment 1 manipulated encoding by requiring 40 young and 40 old adults to process metaphors in one of four conditions: (a) rating them on an imageability, (b) rating them on a semantic relatedness dimension, (c) rating them in terms of number of alternative interpretations, or (d) learning them intentionally. Retention was then measured at three levels of retrieval support: (a) free recall, (b) cued recall (topic and vehicle cues), and (c) recognition, in a within-subjects design. The results indicated that, in the free, recall task, the intentional group recalled more metaphors than the interpretations and imagery conditions, and recall in the three incidental conditions did not differ. Recall improved significantly with increasing levels of retrieval support. Vehicle cues facilitated memory performance more than topic cues. The young remembered more than the old adults at all levels .of retrieval support, and the age by recall task interaction indicated that the difference in their performance was greatest with cued recall, less for free recall, and was minimal with the recognition test. Further, the old improved proportionally more than the young adults across the three retention tasks.
The second experiment comprised two tasks to test the utility of semantic feature, and imagery, models, respectively, with a sample of 20 young and 20 old adults. In the first task, metaphoric noun pairs were presented in or out of sentence context. Participants rated the similarity of the topic and vehicle terms in each noun pair and generated adjectives (features) the two nouns in each pair shared. Free recall of the noun pairs was then tested. The results revealed support for semantic feature theory in the: (a) significant and positive correlation between the rating and feature generation data, and (b) higher ratings and more features produced for high semantic relatedness noun pairs as compared to low semantic relatedness noun pairs. Furthermore, although similarity ratings were not affected by age, the young and old adults generated different shared features for the noun pairs, suggesting that the two age groups may perceive the same degree of similarity between two nouns, but the actual basis of this similarity may be different. The young recalled more noun pairs than the old adults, and the effect of context was not significant in any analysis.
In the second task, high and low imagery metaphors were auditorily presented with or without instructions to image to facilitate memory performance. Free recall, cued recall (topic and vehicle cues), and recognition tests were administered in a within subjects design. The findings demonstrated no effect of imagery instructions on recall. Support for imagery was obtained in the superior recall of high imagery sentences when compared to low imagery sentences and in the interaction between sentence type and instructions. The recall results indicated analogous effects of age, cue type, and recall task to those obtained in the first experiment.
The empirical outcomes of these experiments were discussed with reference to considerations of context, individual differences, the better operational definition of constructs, and the implications for cognitive theories of information processing.
Description
Keywords
UN SDG 4: Quality Education