Training and transfer of clustering and cumulative rehearsal strategies in retarded individuals.

Forty-four retarded adolescents varying in MA, IQ, and etiology were taught one of three active rehearsal strategies or assigned to a control condition. Immediate, short-term (1 day), and long-term (2 weeks) transfer of the strategy was assessed by presenting new 9-item serial lists on each day. Exposure durations, which indicated the rate of subject-pacing, showed that the two highest MA-IQ (mean IQ = 55) groups, one diagnosed as organically impaired and the other as cultural-familially retarded, successfully transferred their strategies over three sessions, while no transfer occurred in the lower MA-IQ group (mean IQ = 40). Greater consistency in the expected pattern of exposure durations of high MA-IQ subjects was found in the cumulative-rehearsal group than in the cumulative-clustering group across sessions. During immediate transfer, the best recall was found for the cumulative group, while on short-term transfer both the cumulative and the cumulative-clustering groups learned more rapidly than the clustering or control groups. No significant recall differences during long-term transfer were apparent for high MA-IQ subjects, probably due to a ceiling effect. For retarded individuals with IQs above 50, the cumulative-rehearsal strategy was the most efficient technique for inducing strategy transfer to a categorized serial list.