The effects of associative strength and response hierarchy on paired-associate learning

Abstract

Summary Two experiments were designed to assess the influence of free-associative strength on the paired-associate learning of children. In each, three groups of fourth-grade children learned lists of ten verbal paired-associates which varied in average associative strength between the stimulus words and the responses. Group I learned a list composed of stimuli and their primary normative responses, five of which were relatively strong and five of which were relatively weak; Group II learned a list composed of the same ten stimuli and responses which occurred to them with an intermediate normative frequency; and Group III learned a list composed of the same stimuli and responses of very low normative frequency. In Experiment I, ease of learning varied with the average associative strength of the pairs at all levels; and differential performance was obtained on Lists I and II between the pairs with stimuli which elicit relatively high-strength primary responses and those which elicit relatively low-strength primary responses. Performance on the former was better than on the latter. Experiment II differed from Experiment I only in that the Ss had a study trial on the lists before learning trials began. The results essentially paralleled those of Experiment I. Such findings have not been obtained with adults and were interpreted as representing differences due to natural language habits of differential strengths. The findings with regard to the differential performance on the two types of pairs were discussed as indications that the forced-frequency nature of the single response normative data caused the relative strengths of the responses following high-strength primary responses to be underestimated when based on absolute normative frequencies.