Evidence for Shared Parenting Domains, Definitions, and Workforce Items: A Confirmatory Factor Analysis

Background: While shared parenting is not a new concept (Landy & Munro, 1998), there has been a renewed interest in shared parenting in child welfare (Klein-Cox, Tobin, & Denby, 2024). Shared parenting has a variety of definitions and alternatives, such as co-parenting, associated with it. However, little is known about the facilitators, barriers, and outcomes of shared parenting. Our lack of knowledge about shared parenting is partly due to inconsistent definitions and evaluators’ inability to measure shared parenting. Thus, to better understand factors (i.e., facilitators, barriers, and outcomes) associated with shared parenting, evaluators will benefit from establishing a definition for and measurement of shared parenting. This study aims to contribute to literature by defining, developing, and validating a measure of shared parenting.

Method: This research is part of the Improving Child Welfare Through Investing in Families (ICWIF) Grant cluster. The purpose of this cluster was to create shared parenting domains and definitions. Guided by theory and literature, ICWIF recipients came to a consensus and defined shared parenting as the intentional practice of developing partner relationships between child welfare staff, caregivers, and parents. ICWIF recipients also identified and created a 25-item instrument for child welfare staff to capture four domains of shared parenting: clear communication (7-items), united front (5-items), normalcy (8-items), and attitudes/philosophy (5-item). Data were collected via a statewide, online survey of child welfare professionals in one Midwestern state (N=439). Evaluators performed a confirmatory factor analysis (CFA) to assess the convergent and divergent validity of the factor structure. Full-information maximum likelihood was used to account for missing data.

Results: Results demonstrate preliminary evidence of reliability and construct validity for each shared parenting domain. After our initial 25-item CFA analysis, we removed items with low primary factor loadings since this is our first attempt to develop these measures. Two communication items and one attitude item were removed. Guided by modification indices and theory, a total of five correlations among error terms were used to improve model fit. The final 22-item model demonstrated convergent and divergent validity of all sub-scales. The final model also showed good fit (χ²=621.03, p=0.00; chi2/df=3.1; RMSEA=0.08; CFI=0.94).

Conclusions: The development and validation of the shared parenting instrument has three important implications. First, the ICWIF cluster established domains and definitions for shared parenting using theory and consensus. Second, this study provides evidence for a new shared parenting measure and validated its internal structure. Third, evidence for the shared parenting measure provides evaluators with a new tool to identify facilitators, barriers, and outcomes related to shared parenting. In sum, as research interest grows about shared parenting, the newly developed shared parenting measure will provide an important tool for evaluators in their future research.