highly significantp < 0.001
There is a much less striking pattern for barriers with men reporting slightly more barriers both in addiction and in recovery and with no gender difference in the reduction in recovery barriers in the time in between. The final analysis attempted to predict the total number of recovery assets in recovery using the following factors as independent variables: age, gender, length of recovery, age of first drug use, age of last drug use, recovery strengths in active addiction, recovery barriers in active addiction and recovery barriers in recovery The final model was highly significant (F = 28.58, p < 0.001), predicting 31.8% of the variance of recovery strengths for individuals in recovery, with only age of first use as a non-significant predictor in the final model (see Table 7 ).