metapoweR

The primary goal of metapower is to compute statistical power for meta-analyses. Currently, metapower has the following functionality:

Computation of statistical power for:

1. Summary main effects sizes
2. Test of homogeneity for between-group variance (for Random-effects models).
3. Test of homogeneity for within-study variance
4. Categorical moderator analyses

metapower can currently handle the following designs and effect sizes:

1. Standardized mean difference: Cohen’s d
2. Correlation between two continuous variables: Correlation Coefficient (via Fisher’s r-to-z transformation)
3. Probability of Success/Failure: Odds Ratio

Installation

And the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("jasonwgriffin/metapower")

Example

library(metapower)
my_power <- mpower(effect_size = .25, sample_size = 20, k = 30, es_type = "d")
print(my_power)
#> 
#>  Estimated Meta-Analytic Power 
#> 
#>  Expected Effect Size:              0.25 
#>  Expected Sample Size (per group):  20 
#>  Expected Number of Studies;        30 
#>  Expected between-study sd:         
#> 
#>  Estimated Power: Main effect 
#> 
#>  Fixed-Effects Model                            0.990698 
#>  Random-Effects Model (Low Heterogenity):       0.962092 
#>  Random-Effects Model (Moderate Heterogenity):  0.8621495 
#>  Random-Effects Model (Large Heterogenity):     0.57799 
#> 
#>  Estimated Power: Test of Homogenity 
#> 
#>  Fixed-Efects Model                             NA 
#>  Random-Effects Model (Low Heterogenity):       0.2926194 
#>  Random-Effects Model (Moderate Heterogenity):  0.9782353 
#>  Random-Effects Model (Large Heterogenity):     1
power_plot(my_power)

See Vignette “Using metapower” for more information

References

All mathematical calculations are derived from L. V. Hedges & Pigott (2004), Bornstein, Hedges, Higgins, & Rothstein (2009), and T. D. Pigott (2012).

Bornstein, M., Hedges, L. V., Higgins, J., & Rothstein, H. (2009). Introduction to meta-analysis. Hoboken, NJ: Wiley.

Hedges, L. V., & Pigott, T. D. (2004). The power of statistical tests for moderators in meta-analysis. Psychological Methods, 9(4), 426–445. https://doi.org/10.1037/1082-989x.9.4.426

Pigott, T. D. (2012). Advances in meta-analysis. NewYork, NY: Springer.

Issues

If you encounter a clear bug, please file a minimal reproducible example on github.