Article Figures & Data
Figures
- Figure 1.
Eligibility and selection of subjects.
- Figure 2.
Flow of subjects through study.
- Figure 3.
Logarithmic change functions for BSI subscales. Time is the average number of months between assessments.
- Figure 4.
Logarithmic change functions for BSI GSI scores for each study group for five assessment periods. Although these lines are not statistically parallel, the interaction term was not statistically significant, indicating that the groups changed at the same basic rate over time. GP, general prison; AS, administrative segregation; MI, mental illness needs; NMI, no identified mental illness needs; SN, special needs prison. Time is the average number of months between assessments.
- Figure 5.
Logarithmic change functions for BSI GSI scores for AS groups for six assessment periods. Although these lines are not statistically parallel, the interaction term was not statistically significant, indicating that the groups changed at the same basic rate over time. AS, administrative segregation; MI, mental illness needs; NMI, no identified mental illness needs. Time is the average number of months between assessments.
Tables
- Table 1
Summary Statistics for BSI Subscales for Raw Scores and Transformed Centered Scores
BSI Scale Assessment Nonpatient Norms Outpatient Norms 0 1 2 3 4 5 M SD M SD M SD M SD M SD M SD M (SD) M (SD) Original Scale Anxiety 1.05 0.99 0.86 0.96 0.87 0.98 0.82 0.94 0.75 0.89 0.62 0.82 0.26 (0.31) 1.51 (0.95) Depression 1.34 1.07 1.12 1.06 1.12 1.10 1.06 1.05 1.02 1.05 0.84 0.91 0.21 (0.33) 1.65 (1.11) Hostility 1.03 0.99 0.84 0.89 0.91 0.95 0.94 1.06 0.88 0.94 0.78 0.94 0.34 (0.40) 1.07 (0.90) Interpersonal sensitivity 1.20 1.05 1.00 0.96 0.99 1.05 0.95 1.02 0.95 1.01 0.78 0.94 0.24 (0.38) 1.48 (1.06) Obsessive-compulsive 1.34 1.01 1.19 1.03 1.18 1.07 1.15 1.05 1.07 0.99 0.92 0.92 0.37 (0.41) 1.53 (0.98) Paranoid ideation 1.52 1.01 1.28 0.98 1.29 1.02 1.26 1.02 1.21 1.02 1.17 0.99 0.33 (0.41) 1.06 (0.93) Phobic anxiety 0.77 0.95 0.62 0.88 0.66 0.92 0.59 0.89 0.56 0.79 0.42 0.80 0.11 (0.25) 0.79 (0.84) Psychoticism 1.28 1.02 1.07 0.95 1.04 0.99 0.97 0.93 0.96 0.93 0.83 0.86 0.15 (0.27) 1.12 (0.84) Somatization 0.82 0.87 0.65 0.75 0.69 0.82 0.62 0.80 0.61 0.79 0.45 0.67 0.23 (0.32) 0.67 (0.71) GSI global score 1.15 0.82 0.96 0.80 0.97 0.86 0.92 0.83 0.88 0.79 0.75 0.72 0.25 (0.24) 1.20 (0.70) Transformed Anxiety 0.00 0.55 −0.15 0.58 −0.13 0.58 −0.16 0.57 −0.21 0.56 −0.29 0.54 Depression 0.00 0.55 −0.14 0.59 −0.16 0.62 −0.18 0.60 −0.21 0.60 −0.31 0.58 Hostility 0.00 0.54 −0.11 0.53 −0.08 0.55 −0.09 0.59 −0.10 0.55 −0.17 0.56 Interpersonal sensitivity 0.00 0.58 −0.11 0.57 −0.16 0.62 −0.16 0.61 −0.17 0.61 −0.27 0.59 Obsessive-compulsive 0.00 0.54 −0.10 0.58 −0.11 0.59 −0.12 0.59 −0.16 0.57 −0.24 0.55 Paranoid ideation 0.00 0.48 −0.13 0.52 −0.13 0.54 −0.15 0.54 −0.19 0.57 −0.21 0.56 Phobic anxiety 0.00 0.59 −0.10 0.57 −0.10 0.60 −0.14 0.58 −0.14 0.55 −0.27 0.53 Psychoticism 0.00 0.53 −0.13 0.56 −0.16 0.57 −0.18 0.56 −0.20 0.57 −0.27 0.54 Somatization 0.00 0.52 −0.13 0.52 −0.10 0.53 −0.16 0.53 −0.17 0.53 −0.26 0.47 GSI global score 0.00 0.41 −0.11 0.44 −0.12 0.46 −0.14 0.45 −0.16 0.46 −0.24 0.44 Time (mo) 0 NA 3.1 .95 6.0 .92 9.1 .92 12.1 .97 15.4 1.37 BSI subscale norms were taken from the BSI manual using male nonpatients and psychiatric outpatients (Ref. 22, p 35). Transformed data were centered at first assessment for entire sample with a square-root transformation to normalize the positively skewed distributions. Time is the mean number of months (with standard deviation) taken from the baseline assessment.
- Table 2
Level 1 Model Statistics for BSI GSI Global Scores to Determine Best Fitting Change Function
Model Estimated −2LLR AIC Cubic polynomial: Int, T1, T2, T3 random* 387.87 417.87 Cubic polynomial: Int, T1, T2 random 387.70 411.70 Cubic polynomial: Int, T1 random 407.92 425.92 Cubic polynomial: Int random 442.20 456.20 Quadratic polynomial: Int, T1, T2 random 391.71 413.71 Quadratic polynomial: Int, T1 random 411.52 427.52 Quadratic polynomial: Int random 445.38 457.38 Logarithmic Model: Int, TLN random 387.68 401.68 Logarithmic Model: Int random 488.72 498.72 Linear: Int, T1 random 425.94 439.94 Linear: Int random 458.39 468.39 Intercept-only model (no change) 521.46 529.46 −2LLR, log likelihood ratio statistic; AIC, Akaike's information criterion: smaller values are better. Int, Intercept coded so it is an estimate of BSI GSI at first assessment period. T1, Time coded for linear function. T2, Time squared coded for quadratic function. T3, Time cubed coded for cubic function. TLN, Time coded for logarithmic function. Time is number of months since initial assessment.
↵* Hessian matrix was positive definite although all convergence criteria were met.
- Table 3
Logarithmic Level 1 Models Testing Group and Group by Time Interactions for All Groups Using 5 Time Periods
Model Statistic Estimates Overall model fit Fit statistic (−2RLL) 290.92 AIC statistic 320.92 Repeated measures error structure Autoregressive variance estimate 0.04 (SE = 0.002, p < .001) Autoregressive covariance estimate 0.64 (SE = 0.24, p = .007) Random coefficients variance matrix Intercept random variance 0.09 (SE = 0.01, p < .001) Intercept-change random covariance −0.002 (SE = 0.003, p = .51) Change parameter random variance 0.01 (SE = 0.002, p < .001) Random Coefficients (Level 1 Parameters) With AS MI Coding AS MI intercept estimate 0.12 (SE = 0.04, p = .007) AS MI change parameter estimate −0.07 (SE = 0.02, p < .001) Comparison of each group from AS MI Intercept AS NMI −0.29 (SE = .08, p < .001) SN MI 0.10 (SE = .06, p = .09) GP NMI −0.48 (SE = .07, p < .001) GP MI −0.08 (SE = .08, p = .26) Comparisons of AS MI with each group on difference in change parameter AS NMI 0.01 (SE = 0.02, p = .60) SN MI 0.02 (SE = 0.02, p = .34) GP NMI 0.002 (SE = 0.03, p = .94) GP MI 0.02 (SE = 0.03, p = .45) Random Coefficients (Level 1 Parameters) With AS NMI Coding AS NMI intercept estimate −0.16 (SE = 0.04, p < .001) AS NMI change parameter estimate −0.06 (SE = 0.02, p = .001) Comparison of each group from AS NMI intercept AS MI 0.29 (SE = 0.06, p < .001) SN MI 0.39 (SE = 0.06, p < .001) GP NMI −0.20 (SE = 0.07, p = .005) GP MI 0.20 (SE = 0.08, p = .008) Comparisons of AS NMI with each group on difference in change parameter AS NMI −0.01 (SE = 0.02, p = .60) SN MI 0.01 (SE = 0.02, p = .68) GP NMI −0.01 (SE = 0.03, p = .69) GP MI 0.01 (SE = 0.03, p = .75) Estimating logarithmic model with five time points using all groups. Models are coded in two ways so that each AS group is compared with all other groups on intercept and change parameters. Model fit statistics and covariance structures are the same for each model. Only the random and fixed parameter estimates change for the two models.
- Table 4
Logarithmic Level 1 Model Testing Group and Group by Time Interactions for AS Groups Using Only Six Time Periods
Model Statistic Estimate Fit statistic (−2RLL) 221.21 AIC statistic 239.21 Repeated-measures error structure Autoregressive variance estimate 0.04 (SE = .003, p < .001) Autoregressive covariance estimate 0.77 (SE = .27, p = .004) Random coefficients variance matrix Intercept random variance 0.11 (SE = .02, p < .001) Intercept-change random covariance −0.009 (SE = .006, p = .12) Change parameter random variance 0.01 (SE = .003, p < .001) Random coefficients (Level 1 parameters) AS NMI intercept estimate −0.16 (SE = .05, p = .001) AS NMI change parameter estimate −0.06 (SE = .02, p = .001) Group intercepts effect (overall) AS MI difference from AS NMI intercept 0.29 (SE = .07, p < .001) Group by time interaction AS MI difference from AS NMI change −0.02 (SE = .02, p = .40) All models (as in Table 2) were re-estimated for the AS groups with six time periods; the logarithmic model was still the best fitting model.
