Understanding visual metaphor: developmental and individual differences

The development of an instrument--the Metaphoric Triads Task (MTT)--for the assessment of metaphoric comprehension is described. In the tradition of earlier cognitive-style research, a visual triad format was adopted that offered three possible pairings of pictorial stimuli, one of which was metaphorical in character. A subject's score reflects the number of metaphoric pairings formed (with appropriate metaphoric explanations) across all of the triads of the task. Data are reported for 12 samples of subjects (ranging from 7 1/2 to 28 years of age) who responded to the MTT in a diverse array of studies. Internal analyses of the MTT yielded satisfactory reliabilities (interjudge and internal consistency) and item-sum correlations. Sex differences were negligible, but progressive improvement in MTT score with age was noted. At the same time, a slight modification of the MTT triad format generated performance levels from younger children that approximated those of children 1--3 years older who had taken the MTT in its standard form. Higher MTT scores were generally obtained by those subjects who attempted more pairings, spent more time at the task, and chose the metaphoric pair as "best" among the alternative pairing possibilities. Correlations of MTT performance with standardized tests of intellective aptitudes and achievements were inconsistent across samples and between the sexes within samples. In contrast, MTT scores were quite consistently related to solving difficult analogies, generating high-quality responses to divergent-thinking tasks, and manifesting broad categorizing and physiognomic sensitivity. Significant correlations between the MTT and a set of verbal metaphoric triads offered convergent validational evidence suggestive of a general metaphoric style. Some relation was found between MTT performance and teacher ratings of figurative language appreciation and esthetic sensitivity, though it appeared that these might be mediated by the teachers' estimate of the child's overall capability. Finally, three experimental training studies were carried out. A requirement of exhaustive pairing and informative feedback on pretest items significantly enhanced the MTT performance of the younger children. The provision of appropriate verbal labels for each picture in a triad also significantly enhanced performance by insuring that children's encoding of the pictures was consistent with the metaphoric linkage in each item.