Development and Validation of a Belonging In Child Welfare Workforce Measure

Background and Purpose: Child welfare workers operate in high-stress, high-stakes environments characterized by heavy caseloads, role strain, and exposure to trauma, contributing to burnout and turnover that undermine service quality and child outcomes. Although workplace belonging is linked to well-being, engagement, and retention, existing measures were not developed for child welfare, and the Belonging Barometer showed inadequate fit in this context. This study reports the development and initial psychometric evaluation of a Belonging in Child Welfare Workforce measure, including item generation, exploratory factor analysis (EFA), and cross-validation of competing four-factor and five-factor models.

Methods: Items were developed in a 2-step process. First, 60 items were developed in conjunction with a Belonging in Child Welfare expert with lived experience in the child welfare system. Next, using qualitative feedback from frontline staff in a Florida child welfare organization and a targeted review of literature on belonging, organizational climate, and child welfare workforce outcomes, items were reduced to 50. During Phase 1 of the project, 50 belonging items were administered to 245 agency employees as part of a multi-year evaluation survey. Data were suitable for factor analysis (Kaiser–Meyer–Olkin = .92; Bartlett’s test of sphericity χ²(1176) = 8251.47, p < .001). An EFA using principal axis factoring with oblique rotation was conducted; items with low communalities or cross-loadings were removed based on statistical and conceptual criteria, and alternative four-factor and five-factor solutions were compared. The four-factor solution yielded eigenvalues of 17.07, 3.69, 2.98, and 2.51 (53.57% variance explained), whereas the five-factor solution added a fifth factor (eigenvalue = 2.00) and accounted for 57.65% of the variance. Communalities for retained items ranged approximately from .35 to .79, with most ≥ .40. In Phase 2, the refined item set is being administered to an independent statewide sample of child welfare workers for confirmatory factor analyses comparing the four-factor and five-factor models and examining convergent validity with constructs such as perceived organizational support and intention to stay.

Results: In the EFA, items formed coherent clusters representing organizational belonging (e.g., transparency, inclusion in decisions, leadership respect), team belonging (e.g., support, collaboration, emotional connection with coworkers), and self/professional identity (e.g., confidence in professional judgment, feeling that one’s work helps people, respect for partner systems), with an additional narrower factor emerging in the five-factor solution. Pattern matrices indicated strong primary loadings (generally ≥ .40) and limited cross-loadings for retained indicators, supporting a multidimensional structure rather than a single global belonging factor. These results informed specification of competing measurement models for statewide validation.

Conclusions and Implications: Findings from the EFA phase suggest that belonging among child welfare workers is best conceptualized as a multidimensional construct with distinct organizational, team, and professional-identity dimensions. Establishing a psychometrically sound Belonging in Child Welfare Workforce measure will equip researchers and agencies with a rigorous tool for assessing belonging and evaluating organizational interventions designed to strengthen climate and retention. A validated, field-specific measure can support measurement-driven improvements in organizational practice and contribute to more stable, supportive environments for workers and the families they serve.