In order to assess S's concept utilization in learning the sorting tasks, preschool children were given either a conceptual sorting task (CST) or a half conceptual sorting task (HST). The results revealed that CST was learned faster than HST by the Ss with high conceptual ability (Exp. I) and in the sorting tasks with high frequency instances (Exp. II), and that both sortings were learned almost at the same rate by the Ss with low ability and in the sorting tasks with low frequency. The superiority of CST over HST is attributable to S's concept utilization in learning the sorting tasks. Whether preschool children can learn in a mediational manner or not depends on S's ability and the nature of the learning task. Kendler, Kendler, and Sanders (1967) assessed the utilization of verbal mediation by the use of a procedure similar to reversal and nonreversal shifts in discrimination learning. College students were required to associate Response A with the instances of one conceptual category (e.g., vegetable) and Response B with those of another (e.g., clothing). On the shift learning, either all instance and response associations were reversed (full reversal shift) or only half of them were reversed (half reversal shift). Kendler, et al. assumed that if Ss use conceptual terms as verbal mediators, full reversal shift would be learned faster than half reversal shift, and that if Ss can not use them, reverse would be true. The result were in line with their assumption. Kendler, Kendler, and Marken (1969) studied these two types of shifts with 4-, 5-, 7-, and 18-year-old subjects.The results were that the youngest Ss learned both shifts almost at the same rate and that full reversal shift was learned increasingly faster than half reversal shift with ages from 5 to 18. They argued that the availability of verbal mediation increased with age, and this was consistent with the verbal mediation theory.Although the procedure used in the above studies appears to be appropriate to assess the utilization of verbal mediation in that they employed verbal materials, there might still remain some apprehensions about their methodology. (a) Training Ss to respond in a conceptual manner may have differential effects on younger and older Ss. It is rather convincing that the training may strengthen conceptual responses in older Ss, but whether this is also true in younger Ss is questionable. (b) What was learned in original learning may not necessarily transfer to shift learning, especially in younger Ss. If this is the case, it is not reasonable to infer the possibility of concept utilization from Ss' performances in shift learning.In order to eliminate these factors, Sugimura (1974) employed the following procedure which involved no shift learning.