Twenty stimulated studies on standardized mean difference and one continuous study characteristic reported by Hox (2002).

1 |

The variables are:

- study
Study number

- yi
Effect size (standardized mean difference)

- vi
Sampling variance of the effect size

- weeks
Duration of the experimental intervention in terms of weeks

Hox, J. J. (2002). *Multilevel analysis: Techniques and applications.* Mahwah, N.J.: Lawrence Erlbaum Associates.

Cheung, M. W.-L. (2008). A model for integrating fixed-, random-, and mixed-effects meta-analyses into structural equation modeling. *Psychological Methods*, **13**, 182-202.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 | ```
## Not run:
data(Hox02)
#### ML estimation method
## Random-effects meta-analysis
summary( meta(y=yi, v=vi, data=Hox02, I2=c("I2q", "I2hm"), intervals.type="LB") )
## Fixed-effects meta-analysis
summary( meta(y=yi, v=vi, data=Hox02, RE.constraints=0,
model.name="Fixed effects model") )
## Mixed-effects meta-analysis with "weeks" as a predictor
## Request likelihood-based CI
summary( meta(y=yi, v=vi, x=weeks, data=Hox02, intervals.type="LB",
model.name="Mixed effects meta analysis with LB CI") )
#### REML estimation method
## Random-effects meta-analysis with REML
summary( VarComp <- reml(y=yi, v=vi, data=Hox02) )
## Extract the variance component
VarComp_REML <- matrix( coef(VarComp), ncol=1, nrow=1 )
## Meta-analysis by treating the variance component as fixed
summary( meta(y=yi, v=vi, data=Hox02, RE.constraints=VarComp_REML) )
## Mixed-effects meta-analysis with "weeks" as a predictor
## Request likelihood-based CI
summary( reml(y=yi, v=vi, x=weeks, intervals.type="LB",
data=Hox02, model.name="REML with LB CI") )
## End(Not run)
``` |

Questions? Problems? Suggestions? Tweet to @rdrrHQ or email at ian@mutexlabs.com.

All documentation is copyright its authors; we didn't write any of that.