VPSPulse Mirrors
High-Performance Open-Source Archive
NEWS
Changes
- For all major functions, the input parameter i2 should now be
specified by the user (i2 = .5 by default)
- All plotting functions plot power curves as a function of model
selection (fixed or random effects)
Update
Changes
- For random-effects models, mpower() now uses a different formula to
account for uncertainty in \(\tau^2\)
(see Jackson & Turner, 2017)
- All plot functions were changes to have a preceding plot_. For
example:
plot_mpower(); plot_homogen_power();
plot_subgroup_power(); plot_mod_power
Additions
- Added subgroup_power(), which computes power to detect differences
in subgroups among studies (i.e., Men vs Women)
- The subgroup_power() has slightly different arguments to allow more
flexibility, especially for Odds Ratio
- Added a plotting function to subgroup_power() called
plot_subgroup_power
- A fully functional shiny application is now available
(https://jason-griffin.shinyapps.io/shiny_metapower/)
New Release
Primary functions
- mpower(): Compute statistical power for meta-analysis
- mod_power(): Compute power for categorical moderator meta-analytic
models
- power_plot(): Visualize a range of power curves
- homogen_power_plot(): visualize a range of power curves for the test
of homogeneity