Psychometric evaluation of a multidimensional family resource assessment tool for schizophrenia (M-FRAT) from caregivers' perspectives in communities of Beijing, China.

Objective: A standardized and comprehensive instrument specifically designed to evaluate the support needs of families affected by schizophrenia is needed. This study aimed to evaluate the psychometric properties of the multidimensional family resource assessment tool for schizophrenia (M-FRAT), which was developed to assess family resources among families affected by schizophrenia from the perspective of family caregivers and to inform family-centered support strategies in primary health care. Methods: A total of 304 family caregivers were purposively recruited from seven community health service centers in Beijing. Validation was conducted in two stages. Stage 1: content validity was assessed by an expert panel using a four-point Likert scale. Stage 2: item analysis, reliability, and construct validity were evaluated using data from a caregiver survey, which included sociodemographic characteristics, the M-FRAT, and the SF-36. Item analysis was performed using critical ratio and item-total correlation analysis. Reliability was assessed using Cronbach's alpha. Construct validity was examined through structural validity and hypothesis testing. Structural validity was evaluated using exploratory factor analysis (EFA) with supplementary principal axis factoring (PAF) and parallel analysis. Hypothesis testing included known-groups validity and associations between M-FRAT and SF-36 scores. Results: The M-FRAT comprises 37 items across seven factors, including three internal family resource domains (financial, psychological-spiritual, and medical treatment-related support) and four external domains (social, cultural, economic-environmental, and medical resources). Item-level content validity was high (I-CVI = 0.88-1.00), and all items demonstrated satisfactory discrimination and item-total correlations. The scale showed excellent internal consistency (Cronbach's α = 0.927). EFA indicated a seven-factor structure explained 70.245% of the total variance, which was further supported by PAF and parallel analysis. The structural support dimension was removed during instrument refinements. Higher M-FRAT scores were positively associated with SF-36 domain scores, providing preliminary support for construct validity through hypothesis testing. Conclusions: The M-FRAT demonstrated good content validity, construct validity, and reliability. It may assist primary health care professionals in evaluating family resources systematically through caregiver reports, which facilitates tailored and family-centered schizophrenia rehabilitation support to reduce the caregivers' burden.