Current social learning theory proposes that motor skill information which is acquired by means of observational strategies is stored cognitively in the form of a conceptual representation. The latter is then said to be used as a basis for response production and as a standard for response corrections which occur as a function of performance feedback. What is not known is how much of skill practice time should be spent in observation of a model as opposed to receiving response-contingent feedback. To answer this question, seven experiemental conditions, which varied in regard to amount of practice time spent either in observing a model or receiving response-contingent feedback, were used with 84 subjects. The subjects were given 60 practice trials on a barrier task prior to performing 20 no-feedback retention trials and 20 no-feedback delayed retention trials. The results of the analysis of the immediate retention effects were somewhat equivocal in terms of providing any definitive answers. Most pronounced in the delayed retention analysis were the results which indicated that four out of five conditions, which included some combination of observational learning and performance feedback, were superior to performance feedback alone. A brief list of references is included. (Author/CB)