Keywords: Family Practice; Psychometrics; Family-Centered Care

Background:

This study aimed to develop and psychometrically validate the Family Involvement in Care Scale-Turkish (FILOPS-T), a novel instrument designed to evaluate family-oriented clinical practices from the patient’s perspective within the family medicine setting.

Research questions:

What is the structural validity and reliability of the FILOPS-T scale?

Method:

A multicenter scale development study was conducted across 88 family medicine units. The study sample comprised 1,400 participants (800 females and 600 males) with a mean age of 44.54 ± 15.76 years. To account for the nested structure of the data and ensure robust estimation, Hubert-White sandwich estimators were utilized in both Exploratory Factor Analysis (EFA, n=707) and Confirmatory Factor Analysis (CFA, n=693). A second-order CFA was performed to evaluate the scale's structural hierarchy and specifically to test whether the multidimensional structure exhibits higher-order unidimensionality.

Results:

The final 41-item scale demonstrated a five-dimensional structure in EFA, with factors showing strong conceptual alignment with the hierarchical stages of family involvement proposed by Doherty and Baird. The subsequent second-order CFA confirmed that these five dimensions load significantly onto a single, global "Family Involvement" latent construct. The model exhibited a good fit to the data (RMSEA=0.055, CFI=0.915, TLI=0.909, and SRMR=0.046). Reliability analysis further supported this structure, with omega values for sub-dimensions ranging from 0.871 to 0.967 and a general factor of 0.922.

Conclusions:

The FILOPS-T is a theoretically grounded and psychometrically robust tool. The confirmation of its unidimensionality through a second-order structure supports the dual utility of the scale; it allows for both detailed assessment of specific dimensions and the calculation of a global index of family involvement, enhancing its applicability for evaluating family-oriented care in primary health services. This work was supported by Çanakkale Onsekiz Mart University The Scientific Research Coordination Unit, Project number: TSA-2025-5231.

Points for discussion:

Our model identified five factors, two of which emerged from the structural decomposition of a single level in Doherty and Baird’s framework, yet Level 5 was not detected. Could the 5th level be realistically measured in this setting, or does its absence confirm the practical boundaries of primary care?