Reliability and Validity of the Mandarin Articulation and Phonological Test: Scale Development and Psychometric Properties for Preschool Children

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Purpose: This study aimed to (a) develop preliminary reference data for the Mandarin Articulation and Phonological Test (MAPT) and (b) examine its reliability and validity for assessing speech abilities in Mandarin-speaking preschool children aged 3–6 years. The study responded to the need for speech assessment tools with systematic psychometric documentation in Mandarin, where standardized instruments with comprehensive scoring systems remain limited. Method: Participants included 148 typically developing children (ages 3;0–6;11 [years;months]) and 76 children with speech sound disorders recruited from preschools and hospitals in Taiwan. The MAPT comprises three subtests: Word (67 items measuring speech ability), Articulation (38 phoneme items measuring articulation ability), and Phonology (13 error pattern items measuring phonological ability). Scale development referenced the Bankson-Bernthal Test of Phonology. Reliability was examined through internal consistency (Cronbach's alpha, Kuder–Richardson 20) and interrater agreement (percent agreement, intraclass correlation coefficient [ICC]). Validity was evaluated through content description, construct identification, and criterion prediction approaches. Results: The MAPT demonstrated good interrater reliability (agreement: .88–.99; ICC = .86–.92, p < .001). Internal consistency coefficients were .95, .88, and .72 for the Word, Articulation, and Phonology subtests, respectively. Content validity was supported by comprehensive phoneme coverage meeting diagnostic standards. Construct validity showed moderate correlations with age ( r = .45–.62) and high intersubtest correlations ( r = .65–.78). Criterion validity revealed significant correlations with percentage of consonants correct (PCC; r = .72–.85) and significant group differences between typically developing children and those with speech sound disorders (Cohen's d = 1.8–2.4), though these estimates should be interpreted with caution given that PCC and group classification were derived from the same MAPT speech sample. Percentile-based descriptive thresholds derived from the normative sample (e.g., scores at or below the 10th percentile) are provided as preliminary benchmarks to inform clinical judgment alongside other assessment data. Conclusions: The MAPT demonstrates adequate psychometric properties for clinical assessment and research applications in Mandarin-speaking populations. The preliminary reference data support age-referenced interpretation of children's speech abilities, with percentile-based thresholds offering descriptive guidance for identifying children who may warrant further evaluation. Supplemental Material: https://doi.org/10.23641/asha.33311883