Assessing Culture and Climate in Prison: Using Community-Engaged Methods to Develop Assessment Tools

Background: Assessment of factors associated with prison climate and culture have largely been developed from an expert viewpoint with minimal, if any, feedback and inclusion of the individuals living and working within correctional settings. Additionally, assessments often conflate culture and climate to the point that distinguishing between the two is cumbersome. Prison reform cannot happen without understanding the prison environment from those who experience it regularly and creating meaningful change in how prisons operate. Here, we discuss a study to develop a survey to capture salient constructs related to the prison environment and conduct initial tests on factor structure using a community-engaged approach.

Methods: This three-phase study was used to produce ways to measure ambiguous concepts important to the culture and climate within a prison setting in one Midwest state. Phase one included literature searching, interviews with incarcerated individuals (n=39) and staff (n=58), and paper questionnaires distributed to a subsample of incarcerated individuals (n=84). Phase two involved pilot testing and cognitive interviews with both groups. Lastly, Phase three used survey data from incarcerated individuals (N=461) and staff (N=149) to conduct exploratory factor analyses to assess factor structure for the identified concepts.

Results: Phase one produced 210 items for an incarcerated persons survey and 225 items for an institutional staff survey. In Phase two, feedback included irrelevance of some items, a need to revise some questions for clarity, and suggestions to shorten the survey. Phase three resulted in scales for nine constructs for staff and eight for incarcerated individuals. Staff scales were reliable α=.70 - .97, except for a construct related to prison reform (α=.65). Scales for the incarcerated population demonstrated lower reliability, though still acceptable (α=.76 - .85), except for contact with loved ones (α=.67), health (α=.65), and culture (α=.57).

Implications: Our work demonstrates the importance of community engagement when producing measures of complex constructs. Prior literature pointed to items that were deemed irrelevant by our community partners and our work identified concepts that were previously un- or under-explored in the literature. Additionally, the feedback from partners was invaluable in targeting the topics most relevant to current prison climate and culture. Future work should include standardized tools but should also leave space for community partner feedback on how the tools capture or miss vital aspects of their experiences.