An evaluation of the psychometric properties of the Purdue Pharmacist Directive Guidance Scale using SPSS and R software packages

Abstract

BackgroundThe Purdue Pharmacist Directive Guidance (PPDG) Scale was developed to assess patients’ perceptions of the level of pharmacist-provided (1) instruction and (2) feedback and goal-setting—2 aspects of pharmaceutical care. Calculations of its psychometric properties stemming from SPSS and R were similar, but distinct differences were apparent. ObjectiveUsing SPSS and R software packages, researchers aimed to examine the construct validity of the PPDG using a higher order factoring procedure; in tandem, McDonald's omega and Cronbach's alpha were calculated as means of reliability analyses. MethodsNinety-nine patients with either type I or type II diabetes, aged 18 years or older, able to read and write English, and who could provide written-informed consent participated in the study. Data were collected in 8 community pharmacies in New Mexico. Using R, (1) a principal axis factor analysis with promax (oblique) rotation was conducted, (2) a Schmid-Leiman transformation was attained, and (3) McDonald's omega and Cronbach's alpha were computed. Using SPSS, subscale findings were validated by conducting a principal axis factor analysis with promax rotation; strict parallels and Cronbach's alpha reliabilities were calculated. ResultsMcDonald's omega and Cronbach's alpha were robust, with coefficients greater than 0.90; principal axis factor analysis with promax rotation revealed construct similarities with an overall general factor emerging from R. ConclusionsFurther subjecting the PPDG to rigorous psychometric testing revealed stronger quantitative support of the overall general factor of directive guidance and subscales of instruction and feedback and goal-setting.