Simple effects tests showed that the experimental subjects increased the number of drawings significantly more than the control subjects did on practiced problems, F(1, 105) = 37.97, p < .01; near transfer problems, F(3, 105) = 29.22, p < .01; far transfer problems, F(3, 105)=42.85, p< .01; and DIFVIS problems, F(3, 105) =7.85, p < .01. However, the experimental group did not make more drawings without considering the nature of the problem categories, F(3, 144) =20.93, p< .01. Compared with the other categories, fewer drawings were generated of DIFVIS problems (see Teble 4). This effect was not apparent in the control condition.