Table 3.
Longitudinal Mixed Effect Models Predicting Psychiatric Symptoms and Employment Issues from Reasons for Choosing Sober Living Houses (SLHs) Included Simultaneously.
| Reasons for choosing SLHs | PDSQ psychiatric symptoms total | ASI employment severity |
|---|---|---|
| IRR (95 % CI LL, UL) | β (95 % CI LL, UL) | |
| Affordability | 0.79 (0.67, 0.93)** | – |
| Wanting to live with others in recovery | – | – |
| Location | – | −0.05 (−0.08, −0.02)*** |
| No place to live/this is my only option | 1.28 (1.04, 1.58)* | 0.02 (−0.01, 0.04) |
| Access to transportation | – | 0.08 (0.05, 0.11)*** |
| Appearance of house | – | – |
| Amenities offered | – | 0.03 (0.00, 0.06) |
| Someone else or a program paying fees | – | 0.05 (0.02, 0.08)** |
Note. Level-1 N = 462; Level-2 sample size n = 48. Final models simultaneously include predictor variables that show P ≤ 0.05 in separate models with the lower limit (LL) and upper limit (UL) for 95 % confidence intervals (CIs) in parentheses. “–” signifies covariate with P > 0.05 in separate exposure variable models in Table 2 and therefore not included in these models.
Values reported are Incident Rate Ratios (IRRs) for negative binomial regression models for the PDSQ model and tobit models for the ASI employment issues model. Time (interview), age, sex, and race/ethnicity were also included in the models, adjusting for random effects of SLHs and within-subjects.
Bold signifies P ≤ 0.05. ***p < 0.001, **p < 0.01, *p < 0.05.