The solution process in children's matching-to-sample

Results indicate that mastery of matching-to-sample problems by 99 kindergartners involved an abrupt shift from chance to perfect performance. Position hypotheses were common before solution, with frequent perseveration. Postcriterion errors were rare. Transfer to new stimuli was usually perfect, regardless of whether the shift was intra- or extradimensional. Latency changes near the point of solution suggested a "solution speed" between that of stereotyped responding and the initial, slow choice times. Comparisons between matching and oddity indicated that oddity was harder and might be mastered by a different process. (24 ref.)