dat.landenberger2005.RdResults from 58 studies on the effectiveness of cognitive-behavioral therapy (CBT) for reducing recidivism in juvenile and adult offenders.
dat.landenberger2005The data frame contains the following columns:
| study | character | (first) author and year |
| pubtype | character | publication type (book chapter, journal article, report, or thesis) |
| country | character | country where study was carried out (Canada, New Zealand, UK, or USA) |
| design | character | study design (matched groups, nonequivalent groups, or randomized trial) |
| program | character | purpose of setting up the CBT program (for demonstration, practice, or research purposes) |
| setting | character | treatment setting (community or prison) |
| designprob | character | indication of study design problems (no, favors the control group, or favors the treatment group) |
| n.ctrl.rec | numeric | number of recidivists in the control group |
| n.ctrl.non | numeric | number of non-recidivists in the control group |
| n.cbt.rec | numeric | number of recidivists in the CBT group |
| n.cbt.non | numeric | number of non-recidivists in the CBT group |
| interval | numeric | recidivism interval (in months) |
| group | numeric | study group (adults or juveniles) |
| age | numeric | mean age of the study group |
| male | numeric | percentage of males in the study group |
| minority | numeric | percentage of minorities in the study group |
| length | numeric | treatment length (in weeks) |
| sessions | numeric | number of CBT sessions per week |
| hrs_week | numeric | treatment hours per week |
| hrs_total | numeric | total hours of treatment |
| cbt.cogskills | character | CBT component: cognitive skills (yes, no) |
| cbt.cogrestruct | character | CBT component: cognitive restructuring (yes, no) |
| cbt.intpprbsolv | character | CBT component: interpersonal problem solving (yes, no) |
| cbt.socskills | character | CBT component: social skills (yes, no) |
| cbt.angerctrl | character | CBT component: anger control (yes, no) |
| cbt.victimimpact | character | CBT component: victim impact (yes, no) |
| cbt.subabuse | character | CBT component: substance abuse (yes, no) |
| cbt.behavmod | character | CBT component: behavior modification (yes, no) |
| cbt.relapseprev | character | CBT component: relapse prevention (yes, no) |
| cbt.moralrsng | character | CBT component: moral reasoning (yes, no) |
| cbt.roletaking | character | CBT component: role taking (yes, no) |
| cbt.other | character | CBT component: other (yes, no) |
Landenberger and Lipsey (2005) conducted a meta-analysis of 58 experimental and quasi-experimental studies of the effects of cognitive-behavioral therapy (CBT) on the recidivism rates of adult and juvenile offenders (see also Lipsey et al., 2007). The present dataset includes the results of these studies and a range of potential moderator variables to identify factors associated with variation in treatment effects.
Personal communication.
Landenberger, N. A., & Lipsey, M. W. (2005). The positive effects of cognitive-behavioral programs for offenders: A meta-analysis of factors associated with effective treatment. Journal of Experimental Criminology, 1, 451–476. https://doi.org/10.1007/s11292-005-3541-7
Lipsey, M. W., Landenberger, N. A., & Wilson, S. J. (2007). Effects of cognitive-behavioral programs for criminal offenders. Campbell Systematic Reviews, 3(1), 1–27. https://doi.org/10.4073/csr.2007.6
psychology, criminology, odds ratios, meta-regression
### copy data into 'dat' and examine data
dat <- dat.landenberger2005
head(dat)
#> study pubtype country design program setting designprob n.ctrl.rec n.ctrl.non
#> 1 Guerra (1990) journal USA rct research prison no 23 29
#> 2 Leeman (1993) journal USA rct demo prison no 15 22
#> 3 Kownacki (1995) chapter USA rct research community no 4 6
#> 4 Larson (1989) chapter USA rct research community no 11 2
#> 5 Finn (1998) report USA rct practice prison no 22 60
#> 6 Goldstein (1989) chapter USA rct demo community no 14 18
#> n.cbt.rec n.cbt.non interval group age male minority length sessions hrs_week hrs_total
#> 1 10 19 12 juveniles 17 50 60 12 1 1.0 12
#> 2 3 17 12 juveniles 16 100 34 26 1 6.0 169
#> 3 2 9 6 adults 28 100 NA 7 1 2.0 14
#> 4 8 5 15 adults 19 100 94 63 1 1.0 63
#> 5 16 66 12 adults 30 100 74 17 5 15.0 240
#> 6 5 28 6 juveniles NA NA NA 13 2 1.5 44
#> cbt.cogskills cbt.cogrestruct cbt.intpprbsolv cbt.socskills cbt.angerctrl cbt.victimimpact
#> 1 yes yes yes no no no
#> 2 yes yes no yes yes no
#> 3 yes no yes yes no no
#> 4 yes no yes yes no no
#> 5 yes yes yes no yes yes
#> 6 yes no no yes yes no
#> cbt.subabuse cbt.behavmod cbt.relapseprev cbt.moralrsng cbt.roletaking cbt.other
#> 1 no no no no yes no
#> 2 no no no yes yes no
#> 3 no no no no yes no
#> 4 no no no no yes yes
#> 5 no no no yes no yes
#> 6 no no no yes yes yes
### load metafor package
library(metafor)
### calculate log odds ratios (for non-recidivism in CBT vs. control groups) and sampling variances
dat <- escalc(measure="OR", ai=n.cbt.non, bi=n.cbt.rec, ci=n.ctrl.non, di=n.ctrl.rec, data=dat)
### fit random-effects model
res <- rma(yi, vi, data=dat)
res
#>
#> Random-Effects Model (k = 58; tau^2 estimator: REML)
#>
#> tau^2 (estimated amount of total heterogeneity): 0.1046 (SE = 0.0352)
#> tau (square root of estimated tau^2 value): 0.3234
#> I^2 (total heterogeneity / total variability): 70.62%
#> H^2 (total variability / sampling variability): 3.40
#>
#> Test for Heterogeneity:
#> Q(df = 57) = 213.6898, p-val < .0001
#>
#> Model Results:
#>
#> estimate se zval pval ci.lb ci.ub
#> 0.4226 0.0605 6.9880 <.0001 0.3041 0.5411 ***
#>
#> ---
#> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#>
### estimated average OR and corresponding 95% CI/PI
predict(res, transf=exp, digits=2)
#>
#> pred ci.lb ci.ub pi.lb pi.ub
#> 1.53 1.36 1.72 0.80 2.91
#>
### examine if number of treatment sessions per week is a potential moderator
res <- rma(yi, vi, mods = ~ sessions, data=dat)
res
#>
#> Mixed-Effects Model (k = 58; tau^2 estimator: REML)
#>
#> tau^2 (estimated amount of residual heterogeneity): 0.0761 (SE = 0.0285)
#> tau (square root of estimated tau^2 value): 0.2758
#> I^2 (residual heterogeneity / unaccounted variability): 62.88%
#> H^2 (unaccounted variability / sampling variability): 2.69
#> R^2 (amount of heterogeneity accounted for): 27.27%
#>
#> Test for Residual Heterogeneity:
#> QE(df = 56) = 146.6400, p-val < .0001
#>
#> Test of Moderators (coefficient 2):
#> QM(df = 1) = 6.0695, p-val = 0.0138
#>
#> Model Results:
#>
#> estimate se zval pval ci.lb ci.ub
#> intrcpt 0.2022 0.1015 1.9931 0.0463 0.0034 0.4010 *
#> sessions 0.0695 0.0282 2.4636 0.0138 0.0142 0.1248 *
#>
#> ---
#> Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
#>
### predicted ORs for 1, 2, 5, or 10 sessions per week
predict(res, newmods=c(1,2,5,10), transf=exp, digits=2)
#>
#> pred ci.lb ci.ub pi.lb pi.ub
#> 1 1.31 1.12 1.53 0.75 2.30
#> 2 1.41 1.25 1.59 0.81 2.45
#> 3 1.73 1.49 2.02 0.99 3.04
#> 4 2.45 1.64 3.66 1.25 4.80
#>