Behavioral and demographic predictors of staff ratings of role model status in a corrections-based therapeutic community for women

Purpose - The purpose of this paper is to examine personal and social network characteristics that predict staff ratings of therapeutic community (TC) resident role model status. Design/methodology/approach - In all, 49 incarcerated female residents tracked interactions with peers, including verbal affirmations and corrections, during a 12-hour period. Two weeks later, staff members were surveyed about their view of participants as role models. Poisson regression was used to analyze resident interactions and demographics as predictors of role model status. Findings - The number of corrections given to peers was positively related to staff ratings of role model status (B = 0.234, SE = 0.088, p = 0.008). The number of affirmations given was negatively related to staff ratings (B = 0.112, SE = 0.051, p = 0.028). Resident phase was positively related to staff ratings (B = 0.256, SE = 0.102, p = 0.012). These values did not significantly change when controlling for affirmations and corrections received from peers, non-programmatic interactions between residents, or resident demographics. Research limitations/implications - These results imply that TC staff judge role model status by resident actions in the community rather than demographics or peer reactions. External validity is limited by the single site, case study design, and the fact that only female TC residents were sampled. Originality/value - This study is the first to track resident peer interactions over the course of a day and to link those interactions to role model status.