Psychometric properties of the Child Oral-care Performance Assessment Scale

Abstract

OBJECTIVES: This study aimed to assess the psychometric properties of Child Oral-care Performance Assessment Scale (COPAS). METHODS: Items for the instrument were developed and pilot tested. This questionnaire was implemented in the Australian National Child Oral Health Study 2012-2014, whose aims included the assessment of oral care performance. This nationally representative sample of 23 538 respondents with complete data was divided into five groups: a main validation group and four cross-validation groups, using blocked randomization. Two scales were constructed, full scale with 37 items (COPAS) and a partial scale with a subset of 31 items (COPAS-Partial). Internal consistency was assessed using Cronbach's alpha. Construct validity was assessed using correlation coefficients, and structural validity was ascertained in the main validation group and confirmed in the cross-validation groups using structural equation models. RESULTS: Cronbach's alpha for COPAS was 0.95, and for COPAS-Partial was 0.94. The convergent validity of global satisfaction with oral care and the subscales was r = 0.29-0.51, and that with the overall scales was r = 0.59 for COPAS and r = 0.59 for COPAS-Partial. COPAS (Root mean squared error of approximation (RMSEA) = 0.06, Comparative fit index (CFI) = 0.90, Tucker-Lewis index (TLI) = 0.89, and Coefficient of determination(COD) = 0.99) and COPAS-Partial (RMSEA = 0.07, CFI = 0.91, TLI = 0.90, COD = 0.97) had adequate fit. Structural invariance was present (P-value = 0.97). CONCLUSION: There was acceptable structural validity, construct validity and internal consistency in the models tested for COPAS and COPAS-Partial. COPAS has potential use in the evaluation of the delivery of dental services to children.