A performance curve for assessing change in percentage of consonants correct-revised (PCC-R)

Purpose: Interperting the rapidly changing speech skills of young children recovering from neurological injury is difficult because developmental expectations are generally available only at relatively lengthy intervals (e.g., 6 or 12 months). In this research note, the authors describe the process of generating a Percentage of Consonants, Correct-Revised (PCC-R; L. D. Shriberg, D. Austin, B. A. Lewis, J. L. McSweeny, & D. L. Wilson, 1997a) performance curve and illustrate some of its applications for assessing change in performance over time. Method: The authors compiled mean PCC-R scores from 16 samples of typically developing children (18-172 months) and used curve fitting to test more than 11,000 statistical models of monthly growth in PCC-R. They selected a parsimonious and 2 developmentally plausible model with R-2 = .9839 (p < .0005) and used it to generate the PCC-R, standard deviation, and standard error expected at each monthly age. Results: The PCC-R performance curve distinguished among 65 children (37-57 months of age) diagnosed independently with normal or disordered speech with a high degree of success. More important, the PCC-R performance curve can be used to identify the points at which children (18-172 months) recovering from neurological injury achieve normal-range consonant production. Conclusion: The curve-fitting approach holds promise as a means of interpreting temporal variations in speech production at a finer grain than existing normative data currently allow.