Summary.TWO experiments were carried out with a total of ninety children aged between 9 and 12 to investigate the effects of verbalising on problem solving. The problems were those of the Progressive Matrices intelligence tests. It was found that performance improved when subjects were asked to give reasons for their choices and these reasons were verbally reinforced. Reinforcement is thought to have had the effect of decreasing stereotyped responses. An analogy between intelligence tests of this type and Harlow's (1949, 1959) ‘learning sets,’ suggested by Cashdan, 1961, is rejec