| Abstract: | Thirty children, ranging in age from 6 to 7 years, were administered a concept of dimensionality test (DT) during which they were required to order sets of stimuli varying in size or brightness along the correct dimension. Training effects were assessed by presenting one-half the Ss with the DT prior to a reversal shift problem and one-half subsequent to it. Ss who scored below the median and made fewer errors on the DT made significantly fewer errors on the reversal problem than did Ss who seored above the median. Moreover, a significantly larger proportion of Ss who scored below the median performed as reversal responders on the reversal problem. The DT showed no significant training effects. The results indicate that responding to the dimension rather than specific values within the dimension underlies the ability to shift solutions to a reversal problem. |
