In order to analyze the relation of the performance and strategies of Ss in discrimination shift learning, homodimensional (HOS) and heterodimensional (HES) shifts were investigated at 3 developmental levels: second grade, fourth grade and junior high school. To identify the strategy Ss took in the course from the original stage to the shift stage, E recorded from the first trial to the tenth the cues which Ss selected on the shift stage and asked Ss how they did on the discrimination task after completing the shift problem. On the basis of these data, strategies of Ss were classified as follows: (1) associative strategies, where Ss learn by the contingency between the response and reinforcement, and (2) conceptual strategies, where Ss utilize a common concept. Results obtained were as follows: (1) HOS was easier than HES. (2) Strategies×ages interaction was singnificant.
2 experiments were conducted to investigate the role of inter-pattern transformation structures in similarity judgements of patterns. The first experiment showed that the similarity ratings of pattern pairs vary according to the transformation structures. The effect of variations appeared more clear-cut for the spatial than for the temporal patterns. In the second experiment, Ss were requested to identify the transformation structures before they judged pattern similarities. It was found that identification of transformation structures is easier for the spatial than for the temporal patterns, and that judged pattern similarities tend to depend more strongly upon the types of transformation structures reported by Ss, than upon those present in the stimuli. These results seem to imply that recognition of transformation structures is the essential factor in the pattern similarity judgements.
72 preschoolers and 72 school children observed the original discrimination of nonreversal shift (NRS) followed by 14 NRS trials by themselves. The observational trials were 10 or 30 in each age group. The correct (_??_) or incorrect (×) responses on the 1st (unchanged pair) and the 2nd (changed pair) trials of NRS were mainly measured. In preschoolers, both dependent observational-learning mode (_??_-× responses: DOL) and independent observational-learning mode (_??_-_??_ and ×-×: IOL) were found, but no observational-trial effects on the modes were found. In school children, DOL occurred significantly more than IOL in both groups, and the observational-trial effect was also found. Subproblem analyses suggested that the _??_-× responses were indeed the “dependent” mode rather than the “independent” one.
In Exp. I, 64 preschoolers were trained with their preferred stimulus dimension in the original discrimination (OL) of optional shift learning (OSL) (Condition D), and another 64 were trained with their nonpreferred one (Condition ND). The Ss in each condition were divided into 2 groups, differing in the amount of training. Under D, OL was learned faster and intradimensional shift (ID) occurred more frequently. Under ND, OL was learned more slowly and ID occurred as frequently as extradimen-sional shift (ED). Under ND, ED decreased but ID increased with overtraining. These findings almost coincided with the predictions from the observing response theory. In Exp. II, the degrees of preference strength of 68 preschoolers were measured, and they were trained in OSL. The effect of strength in dimensional preference upon OSL was found under ND.
The relationship between verbal attitudes and overt behavior was examined in the context of Rokeach's theory. Besides attitudinal variables, social normative beliefs (NBs) and habit (H) were taken into account for prediction of behavior. A questionnaire was administered to 207 freshmen. Attitude-toward-situation (AS) was a better predictor of behavior than attitude-toward-object (AO). Contrary to the theory, a weighted average of AO and AS (AOAS) did not increase the accuracy of prediction beyond AO and AS considered separately. Influence of H was conspicuous, while influence of NBs was weak. The relationship of behavior to predictors was somewhat different from that of behavioral intention (BI); especially H did not predict BI.
The purpose of the present study was to investigate job satisfaction in female workers in self-serviced, discount retailing as compared with middle managements in the same business. A questionnaire, which was also intended to examine the validity of the Herzberg theory, was administered to 80 female workers and 41 middle managements. As a whole, job satisfaction was lower in the female workers than in the middle managements both at the first level analysis (frequencies of pleasant and unpleasant experiences) and at the 2nd level analysis (degrees of satisfaction and dissatisfaction) except for the job security factor as hypothesized. Responses of the middle managements fitted with the Herzberg theory better than those of the female workers. In conclusion, from the personnel management viewpoint, the present study reveals the need for improvement in job contents and job contexts for female workers.