Social facilitation during the initial stage of motor learning: a re-examination of Martens' audience study

This study partially replicated Martens' (1969a) social-facilitation study of motor behavior. His very robust performance findings provided impressive confirmation for Zajonc's hypothesis, and his arousal findings have since been used as evidence for a nonlearned-drive basis for social facilitation. The present study also extended Martens' investigation by examining the separate and combined effects of an audience and videotape camera. The effects due to the presence of the audience and camera were not additive; instead, the audience detrimentally affected subjects' performance consistency and the camera resulted in more trials with errors greater than 30 msec after the performance criteria had been attained. Martens' most robust findings for constant error were not replicated, nor were some of his physiological arousal findings. His pattern of constant error results over all trials is atypical of known learning strategies that subjects use to reduce error over successive trials. Overall, audience effects accounted for only a very small portion of the variance.