A procedure for testing position generalization from articulation training.

Abstract

Generalization was repeatedly tested in individual children during an articulation training program which progressed in phases from isolated sound training to medial position training using nonsense syllables and, if necessary, words. Four children who misarticulated the /s/ in initial, final, and medial positions in words comprised the population of the study. If complete generalization on 12 test probes had not occurred after 3 phases of nonsense syllable training, 3 additional phases (initial, final, and medial position) of word training were administered. Training was terminated when (1) complete generalization had occurred or (2) all phases of the program had been administered. Two children generalized completely at the end of nonsense syllable training, one child generalized after nonsense syllable training, word training, and instructions, and the fourth child did not generalize completely by the end of the study. Although the correctly articulated /s/ began to control the children’s responses in untrained contexts, the results showed that position generalization was not a function of the position trained. As the children progressed through the program, generalization increased, but sometimes failed to stabilize on specific items from phase to phase.