No abstract

No abstract

Two paired associate lists of 32 nouns were presented to subjects in three treatment conditions: One received imagery instructions, another saw provided pictorial mediators, and the third received only task instructions. Concrete high-imagery-arousing nouns were superior in facilitating learning, especially in the stimulus position. This effect was reduced in the provided-mediator group, as evidenced by an interaction between pair type and treatment, but it was not eliminated. A post task test of mediator recall presented to the provided-mediator group showed no difference in recall due to pair type if the JIlediator was recalled, but there were differences in recall of the mediator related to.concreteness. The difference between subject-generated and experimenter-provided mediators appeared to depend on the concreteness of materials.It is well established that concreteness (C) and imagery value (I) of words are highly related to recall in paired associate learning (PAL) tasks, especially on the stimulus side, with difficulty increasing in the order: concrete-concrete, concrete-abstract, abstract-concrete, and abstract-abstract (paivio, 1965, 1971). One explanation for this order is that it is easier to form mediatorial images with concrete words than with abstract words. If so, one should be able to reduce the difference between concrete and abstract words by providing a visual mediator for both types of word, thus eliminating the difference in difficulty of producing a mediator. TIlis expectation was not supported in an experiment by Wollen and Lowry (1971) in which they found no interaction between word imagery and provision of visual mediators, but this lack of support may be specific to the procedures of their study. Silhouettes depicting the pair were shown after the stimulus word, but before the response word of the pair had been shown. The task of relating the specific terms of the provided mediator to the more general terms of the low-I pair may have been rendered somewhat more difficult by this procedure, so that mediators were not as effective for low-I pairs as for high-I pairs.The present study explored the relationship between imagery use and concreteness by comparing recall of the four pair types by three groups: (1) a group seeing provided mediators along with presentation of the complete pair, (2) a group instructed to image, and (3) a group given only traditional PAL instructions. With these conditions the usual concreteness effect was expected in the imagery-instruction group and in the control group, but this effect was predicted to be much smaller in the provided-mediator group, because the Requests for reprints should be sent to Frank W. Wicker,

No abstract