Social reinforcement as a function of Task Instructions, sex of S, age of S, and baseline performance

Abstract

Forty boys and forty girls at each of two age levels (Grades 1 and 2 vs Grades 4 and 5) performed at a marble-dropping task, which was introduced as either a game or a test. Four females served as Es. The Es were nonresponsive to S during a baseline minute, and during the subsequent 5 minutes of the task either made supportive comments (Social Reinforcement Condition) or continued to play a neutral role (Nonreinforcement Condition). Within each age level, Ss were blocked according to the sample's median base rate level of response. and base rate was crossed with the other variables in the analysis of difference scores. This procedure minimized group differences in baseline performance, a desirable control in the analysis of difference scores. For older boys, social reinforcement resulted in higher difference scores under game than test instructions, whereas nonreinforcement produced higher difference scores for test than game instructions. For older girls, an opposite pattern of results was obtained, resulting in the significant triple interaction between Sex of S, Reinforcement Condition, and Task Instructions for older Ss. For younger Ss, social reinforcement produced higher difference scores in general than did nonreinforcement. At both age levels, base rate level of S did not interact with any of the major findings. Base rate level of response was, however, a highly significant general determinant of change in rate of response, with both older and younger Ss of high base rate level showing steady decreases in performance across the subsequent 5 minutes of the task. The results were related to previous findings and the advantages of blocking on baseline in analyses of change were discussed in general and in light of the present findings.