Three extrinsic factors were manipulated in a problem-solving task performed by 80 ten-year-old children of I Q 90-110. The factors were: the presence or absence of the name of the concept given in feedback, specific instructions, and whether the child was allowed to verbalize. The task was a modification of Vygotsky's experiment on concept formation using concrete material. The performance criteria were: time taken to solution, number of moves, and amount of verbalization. The presence of the name of the concept given in feedback speeded the time to solution and reduced the number of moves required. Specific instruction speeded the solution but did not reduce the number of moves, whereas verbalization reduced the number of moves but increased the time of solution. (Author)