本研究の目的は, 1)弁別移行学習の分野へ実験教育を導入して,正刺激の値名をいかにして見つけ出すかという解決ストラテジーについての行為を幼児に教え,同時に, 2) Гальперинらの"知的行為の多段階形成説"の第2段階や第3段階の意義を,この分野において検討することであった。 被験児は,4〜6歳児104名。2×2×4の要因配置--言語による認識能力の2水準(上位・下位),RSとNRSの移行型,教授法I〜IIIと統制群IV--が用いられた。事前テストにおける次元間比較課題の成績で被験児を2水準に分け,その後1週間を経て,2次元2値による弁別移行学習を行った。実験教育は,先行,移行の両学習の間に実施された。その1か月後,フォローアップ・テスト(把持テストと転移テスト)が行われた。 主な結果は次の通りであった。1)教授法IとIIは,上位・下位の両群に対して有意に効果があった。2)教授法IIIは,上位群には有意に効果があったが下位群にはなかった。3)教授法Iは,持続性と転移可能性という点でも効果があった。4)教授法IIは,転移可能性の点で有意な効果はなかったが,持続性という点では効果がないとは必ずしも言えなかった。

The purposes of the present study are, 1) to introduce experimental education into the field of discrimination shift learning, and to teach a strategy for discrimination learning so that infants and young children can be mediational learners, 2) to examine the meanings of the 2nd stage (in which external acts based on objects are formed) and the 3rd stage (in which acts on the level of outer speech are formed) in P. Ya. Gal'perin's "theory of stage by stage formation of mental actions", on which the present teaching programs are based. One hundred and four 4-6 year-old preschool children served as subjects (Ss). A 2×2×4 factorial design was used which incorporated two levels of verbal ability (higher group: H. G. and lower group: L. G.), reversal shift (RS) and non-reversal shift (NRS) learning, 3 types of teaching methods (I-III) and a control group (IV). In the pre-test, Ss were classified into two levels according to the scores for the task which required Ss to compare the sizes of interdimensional value-classes. This task is analogous to Piaget's "class inclusion task" (K. Shibata, 1976). Discrimination shift learning was crraied out one week after the pre-test. The problem in this learning was of two dimensions. The criterion at first and second learning was 10 consecutive correct responses. The experimental education was undertaken between the first and second learning. Teaching type I was that of controlling both external acts based on objects and acts on the level of outer speech. Type II was that of controlling only the external acts based on objects. Type III was, that of controlling the acts on the level of outer speech only. The Ss'acts in control group IV were not controlled by the experimenter on either stage. About one month after this experimental education, a follow-up test was given consisting of a retention test and a generalization test. The main results were as follows: 1) Teaching type I and II had a significant effect not only on the H. G. but also on the L. G. 2) Type III had a significant effect on the H. G., but not on the L. G. 3) Type I had a significant effect in terms of retention and generalization. 4) Type II had an effect in terms of retention, but had no significant effect in terms of generalization. Discussion: 1) The 2 criterions which have been frequently used to judge the effect of verbalization (or the occurrence of mediational responses) may not be useful. (a) There is a higher increasing in number of RS learners in the optional shift learning as compared to NRS, b) The RS task is more easily learned than the NRS task.) If the S responds logically in either RS or NRS learning with the 2 dimensional (2 values) stimuli, he could attain the criterion in a few trials, and hence there should be no significant difference between the two shifts. 2) The strategy, which would be used "in his mind" in the case of the human adult, "was externalized" to be taught to younger children. By doing this, the children could develop their strategy in a visible form, and the experimenter could control their thinking on that level. 3) It can be said that the Ss achieve the level of generalization when their acts of strategy shift from the stage of external acts based on objects to the stage of outer speech. For in shifting to the verbal stage they may free their acts from a direct dependence on concrete objects.