Correlation between the measure for cluster proximity (MCP) and the percentage of consonants correct (PCC)

Abstract

ABSTRACT A measure was recently proposed for the quantitative evaluation of consonant clusters in child developmental speech, the measure for cluster proximity (MCP). Its advantage over the widely used measure of the percentage of consonants correct (PCC) is that it differentiates the three main stages of cluster production: reduction, vowel epenthesis, and two-members, correct or not. The purpose of the present article is two-fold: a) to contrast MCP and PCC in two-member word-initial clusters produced within children with speech sound disorders (SSDs), tested before and after therapy, as well as in running speech within typically developing children, and b) to examine the correlation between the two measures across children in typical or atypical speech development. Data used are comprised of 34 sets with 18 word-initial cluster types, taken from the speech samples of 18 children. It is found that while there are generally large deviations between the two measures in individual clusters within and between children, the measures have a strong and statistically significant correlation when averaged over all clusters per child, or over all children per cluster class based on the manner of articulation. For the data at hand, a linear equation is obtained yielding MCP values from PCC values with a high coefficient of determination. The results suggest that PCC norms that are used to assess consonants can be used to predict MCP norms in order to assess consonant clusters in child developmental speech.