Abstract
This study intended to develop and assess the psychometric properties of a caring behavior scale on healthcare students and providers (CBS-HSP) in a Taiwanese population. After a literature review was conducted and an expert panel was consulted for item generation, 104 items for the CBS-HSP scale were derived on a nine-point Likert scale, with 9 indicating 'extremely important' and 1 indicating 'extremely unimportant.' A pilot study was then conducted with seven hundred forty-eight healthcare students and providers in Taiwan for further data analysis. The statistic software used in the study was SPSS for the exploratory factor analysis (EFA) and AMOS for the confirmatory factor analysis (CFA). Also, to examine the psychometric properties of the scale, internal consistency, convergent validities, discriminant validities, and model fit indices were calculated in the study. The EFA results derived 31 items in four factors, with 65.742% of the total variance explained: 'support and attentiveness' (11 items; 48.714% of the variance explained), 'professional knowledge and skills' (8 items; 8.226% of the variance explained), 'gratifying needs and responsiveness' (7 items; 5.236% of the variance explained), and 'confidentiality and trust' (5 items; 3.566% of the variance explained). The Cronbach's alphas for the four subscales and the overall scale ranged from 0.894 to 0.964. The CFA results yielded the same 31 items, with the same four factors. The CFA results demonstrated good to excellent model fit in the χ2/df ratio (1.242), GFI (0.988), CFI (0.988), TFI (0.985), and RMSEA (0.031). The Cronbach's alphas ranged between 0.866 and 0.971; the composite alphas ranged between 0.854 and 0.964. The convergent and discriminant validities also proved the stability of the CBS-HSP scale. The research results indicated that the developed CBS-HSP appeared to be a reliable instrument to measure healthcare students' and providers' caring behaviors.