モデルの遂行水準, 代理自己強化, 代理外的強化が観察者の遂行と自己評価様式に及ぼす効果

抄録

The present experiment was designed,(1) to clarify the functional characteristics of vicarious self reinforcement comparing with the function of vicarious external reinforcement, and (2) to estimate the relative effect of modeled performance level and modeled vicarious reinforcement patterns to the performances and self evaluative pattern of observers.
The subject were 86 boys and 72 girls, ranging in age from 10 years 5 months to 11 and half years, drown from 5 classes in 5th grade in a public primary school. Model was a 19 years old female undergraduate, and experimenter was a 23-year old male graduate.
So as to equalize the task abilities in each experimental and each control groups, the figurenumber substitution task was first administered to all subjects in classroom settings. Then, model and/or each subject was administered three trials of simple arithmetic calculation based upon figurenumber substitution (e. g. _??_) in a booklet. 18 calculation problems were printed, for each trials, Model and subject were given 50 sec. to resolve the problems and 30 sec. inter-trial intervals to rest and to evaluate one's performance. The self evaluation of each subject was measured on the 3 point-graphic scale (not enough, nough, highly enough) after each trial.
A 2×2×2×2 factorial design was employed. Half of the experimental children observed high performance model, and half were exposed to low performance model. In both conditions, half the children observed vicarious reward. While the remaining children in each group witnessed vicarious punishments. In addition, each group was further divided into external (reinforcement was administered by experimenter) or internal (model's self reinforcement) control condition.
In high performance condition, model performed 10 problems, and in low performance condition, model performed 6 or 7 problems in each trial. These scores were determined based upon the performance level of no-model control group (Mean=8.25, SD=1.87) of 39 subjects.
In reward condition, model or experimenter evaluated model's performance as highly enough, and in punishment condition, they devaluated model's performance as not enough.
The main results were p resented in FIG. 1. 2. 3. Analysis of variance performed on over-all self evaluation scores disclosed that highly significant effects were produced by the vicarious reward punishment (F=52.85, P<0.001), and not produced by the other variables. So, vicarious self reinforcement and vicarious external reinforcement were found to be equally efficacious in the modification of self evaluative behavior of observers.
Children who obsarved model's self punishment raised significantly their evaluation more affirmatively from first trial to third trial.(t=3.16, P<0.01) But children who observed model receive external punishment did not altered their evaluative standard. So we consider that there should be subtle but important differences between the informative functions of vicarious self reinforcement and vicarious external reinforcement. Vicarious self reinforcement may transmit the self-monitoring flexible self reinforcement pattern, while vicarious external reinforcement, may transmit other-oriented inflexible self reinforcement pattern.
Analysis of variance on performance scores revealed that any independent variables we introduced did not produce significant differences.