Abstract
Purpose. The purposes of this investigation were; to determine if there were any differences between elementary-aged field independent (FI) subjects and field dependent (FD) subjects in their respective abilities to generate and recall inferences; to determine if they generated and recalled inferential sentences with greater efficiency than premise, verbatim sentences; and whether the performance of the FI Ss would be in accordance with the semantic integration theory (superior inference recall) and if the performance of the FD Ss would be in accordance with the semantic transformation theory (premise sentence recall).Procedures. One hundred and ten fifth and sixth grade participants in a remedial summer school program were administered the Group Embedded Figures Test. Based on a quartile splitting a stratified, random sample of fifty-four Ss were identified; eighteen FI, eighteen middle range, and eighteen FD.Each S participated in three randomly sequenced activities that contrasted inferential and verbatim memory and recall. In the first activity one FI, one MR, and one FD Ss were read seven paragraphs, each containing three premise sentences followed by twenty-eight recognition sentences which tested premise and inference sentence memory accuracy. More complex paragraphs containing multiple types of inferences and verbatim statements were presented in the second activity. Four different types of inferences and two types of verbatim statements were tested. The third activity involved the reading of seven short paragraphs containing four-element linear orderings (A B C D). Following each presentation the Ss responded to twelve true-false questions concerning the inferential remote and the verbatim adjacent pairs.Eleven directional hypotheses at the .05 level of significance tested which theory of memory was utilized, predicted superior inference generation by the FI Ss, and predicted superior verbatim sentence recall by the FD Ss. Analysis of Variance and the "t" test for independent and correlated samples were used.Findings. All four hypotheses dealing with the contrasting theories of memory produced significant findings. Ss from all groups on all activities recalled inferences and inferential relations with greater accuracy than premise, verbatim sentences and inferences that were a composite of several events or a summary of several ideas were remembered with greater accuracy than inferences that contained one idea or event.On the first and third activities the FI Ss recalled significantly more inferences and inferential remote pairs than the FD Ss. On the second activity the FD and FI Ss were equally proficient on both types of sentences. In only one instance did the FD Ss perform in accordance with the predictional hypotheses. Contrary to prediction, the FD Ss did not recall significantly more verbatim sentences than the FI.Conclusions. (1) The semantic integration theory of memory was consistently supported by the performance of the Ss on all activities. (2) Fifth and sixth grade children automatically create inferences during the inputting/perceptual stage when presented with premise sentences and remember these inferences more efficiently than the premise sentences. Inference generation did not occur during the recall/retrieval stage. (4) In contrast to the FD, the perceptual style of the FI can now be described by the additional characteristics: (a) A greater efficiency in inference generation and recall. (b) The utilization of a memory strategy that involves the semantic integration of inferential summative information rather than the integration of premise, verbatim sentences. (c) A generally superior memory efficiency.