Short-form of the Parental-Caregivers Perceptions Questionnaire: psychometric properties of the Mexican version

Abstract

Objective. To explore the psychometric performance of the Parent-Caregiver Perceptions Questionnaire short-form in Mexican parents. Methods. Cross-sectional study performed in a sample of 214 participants. The content validity of the P-CPQ-16 and two alternative 14-item scales was examined using exploratory factor analysis. The goodness-of-fit measures of three models were calculated in the confirmatory factor analysis. Reliability measures (Cronbach’s alpha and McDonald’s omega) were calculated. Internal consistency and discriminant power were evaluated using Spearman’s coefficient and success percentages. Results. The P-CPQ-16 with four factors explained 57.88% of the variance. However, the items ‘slow eating’ and ‘shy or embarrassed’ were not satisfactory. The P-CPQ-14 three-factor scale explained 50.22% of the variance with better interpretability of the structure compared to the P-CPQ-16. Reliability of P-CPQ-16 was acceptable (alpha=0.70; omega=0.70) as well as the trifactorial P-CPQ-14 (alpha=0.71; omega=0.72). Confirmatory analysis between models revealed similar goodness-of-fit indicators: model (16 items-4 factors): X2=43.765 (df 62) p=[0.961], GFI=0.98, CFI=0.98, RMSEA=0.036, TLI=0.96, AIC=508.26, model (14 items-4 factors): X2=24.43 (df 41) [p=0.981], GFI=0.99, CFI=0.99, RMSEA=0.029, TLI=0.98, AIC=423.702, model (14 items-3 factors): X2=40.183 (df 52) [p=0.883], GFI=0.98, CFI=0.98, RMSEA=0.035, TLI=0.97, AIC=365.38. Conclusions. The three versions of the P-CPQ showed an acceptable psychometric performance. The three-factor version P-CPQ-14 presented better structure, high factor loads, good model fit, adequate parsimony, and theoretical plausibility. This alternative version needs to be psychometrically tested on a larger sample.