Assessment of Wellness Knowledge of Prisoners in a State Correctional System Wellness and Fitness Program

Background: Prison populations face the growing problems of decreasing fitness, drug addiction, poor mental health, and violence. Wellness programs can help counter these strains on the correctional system and the overall well-being of inmates. These programs help yield a lower rate of recidivism, better prisoner attitudes, and a higher likelihood of inmate contribution to society upon release. Purpose: This study assesses whether a prisoner wellness program test adequately measures wellness knowledge, shows validity as a measure of wellness knowledge, and serves as a good measure for wellness course participants versus non-participants. Methods: Prisoner responses to the dichotomous 25 item multiple choice state corrections department administered Wellness Education Exam developed for a wellness program was evaluated using Item Response Theory (IRT) models, specifically two-parameter loglikelihood (2PL) and three-parameter loglikelihood (3PL) models. Results: Unidimensionality was determined through a principal component analysis. Fit statistics were obtained for the 2PL and 3PL. The 2PL model yielded the following: AIC of 3739.16, BIC of 3893.23, and a loglikelihood of -1819.58. The 3PL model yielded AIC of 3757.50, BIC of 3988.60, and loglikelihood of -1803.75. The Log ratio test was 31.67, df=25 and a p-value of 0.168. Thus, the 2PL model had better fit. Also, participants in the wellness education program performed better on the exam than those in the control group. Conclusions: Findings are consistent with previous literature that prisoners with access to wellness education have health and exercise knowledge. Results suggest minor revision to the test as well as re-administration of the wellness program.