Effects of Listener Training on Differential Recognition of Monosyllables: Implication for Audiology and Intelligibility Testing

Abstract

The effects of spaced training upon listeners' ability to recognize monosyllables in noise are considered in the light of previous studies and the predictions of information theory. Twenty-four American male listeners were alternately trained and tested at weekly intervals over several months on a sample of 600 English monosyllables which phonetically agglutinate on a given base. Differential recognition of these words according to the component number of speech sounds was charted. Correlations of stage-by-stage results with frequency of linguistic occurrence and number of words available within sound class are presented. Articulation scores of groups and individuals are shown to vary markedly on test and retest as a result of training and the number of phones in a test word. Both training and phonemic length of test words are seen as primary variables, optimum levels of which should be defined for all phases of audiological and intelligibility testing. The above work was carried out under contract with Operational Applications Laboratory, Washington, D. C.