Effect Size for Single-Subject Design in Phonological Treatment

The purpose of this study was to document, validate, and corroborate effect size (ES) for single-subject design in treatment of children with functional phonological disorders; to evaluate potential child-specific contributing variables relative to ES; and to establish benchmarks for interpretation of ES for the population.

Data were extracted from the Developmental Phonologies Archive for 135 preschool children with phonological disorders who previously participated in single-subject experimental treatment studies. Standard mean difference(all with correction for continuity) was computed to gauge the magnitude of generalization gain that accrued longitudinally from treatment for each child with the data aggregated for purposes of statistical analyses.

ES ranged from 0.09 to 27.83 for the study population. ES was positively correlated with conventional measures of phonological learning and visual inspection of learning data on the basis of procedures standard to single-subject design. ES was linked to children's performance on diagnostic assessments of phonology but not other demographic characteristics or related linguistic skills and nonlinguistic skills. Benchmarks for interpretation of ES were estimated as 1.4, 3.6, and 10.1 for small, medium, and large learning effects, respectively.

Findings have utility for single-subject research and translation of research to evidence-based practice for children with phonological disorders.