Provider Trust at the Intersection of Race, Region, and Supervision: Scale Development and Psychometric Evaluation Among People Under Community Supervision In the U.S. Deep South

Background: Provider trust is a well-documented determinant of health and service engagement, yet valid and reliable instruments for measuring trust among people in community supervision programs (e.g., probation, parole) do not exist. Provider mis- and distrust among this population drives disengagement from health care, mental health services, and community supervision programming — compounding structural inequities in health and reentry outcomes. Without measurement, the quality of those provider relationships cannot be examined or improved. To our knowledge, this is the first study to psychometrically evaluate a provider trust instrument for people under community supervision, to develop and introduce a trust scale informed by the lived experiences of Black women under community supervision in the Deep South, and to empirically establish the association between provider trust and mental health outcomes in this population.

Methods: Survey data were collected across eight community supervision sites in Alabama from two parent studies (combined N = 439). The sample was predominantly male (78%), equally Black or African American and White (37.6% each), largely Alabama-born (79.9%), and between the ages of 30 and 59. Participants carried substantial trauma burden: mean ACE score 3.65 (SD = 2.61), mean PCL-5 score 29.13 (SD = 17.34), and mean CES-D score 16.71 (SD = 10.79). Provider trust was measured using a 12-item subscale adapted from the Working Alliance Inventory (WAI; n = 54). A 24-item Southern Specific Trust Factors scale (SSTF), developed from prior qualitative interviews with Black women under community supervision, was administered to the Black female subsample (n = 53; eligibility required a trusted provider relationship of at least five years) and completed by n = 16. WAI and SSTF evaluation included Spearman correlations, linear regression, exploratory factor analysis, and Kruskal-Wallis tests. Mokken scale analysis in R 4.5.2 evaluated SSTF scalability via Loevinger's H coefficients at the item, item-pair, and scale levels.

Results: The WAI subscale demonstrated strong psychometric properties, with a two-dimensional structure distinguishing positive relational trust from fear-based mistrust and inter-item correlations ranging from ρ = 0.50 to ρ = 0.89. Higher trust was associated with lower depressive symptomatology (ρ = -0.469, p < .001) and lower PTSD symptom severity (ρ = -0.387, p = .004), while higher ACE scores were associated with greater provider discomfort (ρ = 0.324, p = .04), indicating that distrust is rooted in cumulative trauma exposure. Kruskal-Wallis testing identified significant group differences in trust patterns by residence, food and shelter access, race, religion, and mental health experiences. Among the Black female subsample, Mokken analysis of the SSTF yielded preliminary scalability coefficients supporting item retention, interpreted as evidence for larger-scale validation.

Conclusions and Implications: Provider trust is a clinically significant variable: its correlates are measurable and improving it has concrete implications for health outcomes. Notably, eligibility to complete the SSTF was met by only 16 of 53 Black women — underscoring the depth of provider mistrust in this population and the need for research and practice capable of reaching those whom systems have most consistently failed.

Keywords: community supervision, psychometrics, Working Alliance Inventory