RESUMO
Allowing learners to control feedback has been an effective strategy in motor skills learning. However, most studies of self-controlled (SC) feedback have used simple tasks that may be dissimilar to sports skills that generally demand more degrees of freedom and cognition. Thus, this study investigated the effects of SC knowledge of results (KR) on learning a complex Taekwondo skill. Twenty-four undergraduate volunteers of both sexes, aged 18-35 years, practiced a specific serial Taekwondo skill that was novel to them. We divided participants randomly into SC and yoked groups and compared their performance after they learned a specific displacement sequence, finishing with a lateral kick (bandal-tchagui) at a punching bag within a target time span. During acquisition, all participants performed 48 trials divided into six blocks and, on a retention test 24 hours later, they performed 10 more trials. We found that both groups reduced their errors from the first to the last block of the acquisition phase and that the SC group showed a better performance on the retention test, relative to the yoked control group. SC KR participants requested KR mainly after good trials, though they showed no statistically significant differences between trials with and without KR. Their inefficiency in estimating their own errors may have been due to task complexity, since many aspects of the task beyond its temporal requirement demanded the learners' attention. Our results, using a novel Taekwondo serial skill, confirm and extend the benefits of SC KR from just simple motor learning in past studies to learning complex motor skills.