PERSISTENCE OF GAINS FROM INSTRUCTION IN CLASSIFICATION, SERIATION, AND CONSERVATION

Children who received learning-set instruction on unidimensional class ification, unidimensional seriation, and number conservation in the mi ddle of kindergarten were retested at the end of 1st grade. The follow -up tests showed that a 17-percentile-point advantage conveyed by this instruction on the O-LSAT measure of school-related reasoning was ess entially intact. It also appeared that a 10-percentile-point gain in m athematics, presumably an indirect effect of the increased competence in reasoning, was also intact. On the other hand, there was no longer any difference between experimental and control children in verbal com prehension. These results are discussed in terms of temporary, permane nt, and self-propagating gains.