A Phonemic Analysis of Half-List Speech Discrimination Tests

Publisher Website
Google Scholar
Abstract
The results of a phonemic analysis of the words in the W-22 test and PB test lists 7, 8, 9, and 10 are reported. The question asked was: what happens to phonetic balance when a 50-word test is split into two halves? The method involved phonemic transcriptions of the word lists from which frequency distributions of the phonemes were made. Pearson product moment correlation coefficients were computed using the frequency distribution data for the first half versus the second half of each list and for each half-list versus the frequency of occurrence reported by Dewey. The results of the analysis showed that the highest r between split halves of a test was .57 and the lowest .17 and that rs for PB and W-22 tests were similar. Correlations between half-lists and the Dewey data ranged from .19 to .90 with the coefficients being somewhat poorer for the PB test than for the W-22 test. Some observations regarding phonemic distribution in the two tests were made. The analysis showed that the PB characteristic of the whole test is lost when a 50-word list is split into two 25-word lists.

This publication has 0 references indexed in Scilit: