Adapting and Validating the Client Engagement Scale for the Danish Child and Family Welfare Context: A Rasch-based construct validity study

Background: Client engagement encompasses family involvement in child protection casework and the collaborative relationship between case workers and families. The Client Engagement Scale was developed to assess
client engagement within child protective systems.
Objective: Building on Yatchmenoff’s conceptual framework, this study aimed to translate and conduct an initial
psychometric validation of a client engagement measure specifically within Denmark’s child and family welfare
system.
Participants and setting: Parents with active cases in the Danish Child and Family Welfare system completed 210
online survey questionnaires.
Methods: The original pool of 37 items was translated and adapted to the Danish context. Construct validity for
five subscales of client engagement was examined using Rasch models (RM) and graphical loglinear Rasch
models (GLLRM) with a focus on differential item functioning and local dependency of items.
Results: After item elimination, a 5-item expectancy subscale fitted the pure RM. A 6-item receptivity, 4-item
mistrust, and a 3-item working relationship subscale presented some differential functioning and/or local dependency, but to a degree which could be accounted for using GLLRM. The investment subscale did not fit the
RM nor a GLLRM. The resulting scales demonstrated high to excellent reliability, but poor targeting to the study
population.
Conclusion: Essentially valid subscales resulted for four out of five aspects of client engagement within Denmark’s
Child and Family Welfare system. While these subscales show promise for statistical use, future research should
examine their psychometric properties in larger and more diverse samples before attempting to use them at an
individual level in an everyday child and family welfare practice.