powerlmm: Power Analysis for Longitudinal Multilevel Models
In powerlmm: Power Analysis for Longitudinal Multilevel Models

Description Details Supported models Tutorials Author(s) See Also

The powerlmm package provides a fast and flexible way to calculate power for two- and three-level multilevel models with missing data. The focus is on power analysis for the test of the treatment effect in longitudinally clustered designs, i.e. where the first level is measurements, and the second level is subjects nested within a (optional) higher level-three unit, e.g. therapists.

All study designs are specified using the function study_parameters, which lets you define your model using familiar notation, either by specifying the model parameters directly, or by using relative standardized inputs (e.g. % variance at each level). Several functions are provided to help you visualize and understand the implied model, type methods(class="plcp") to see available methods. The basic features of the package are also available via an interactive (Shiny) web application, which you can launch by typing shiny_powerlmm().

The purpose of powerlmm is to help design longitudinal treatment studies, with or without higher-level clustering (e.g. by therapists, groups, or physicians), and missing data. The main features of the package are:

Longitudinal Two- and three-level (nested) linear mixed models, and partially nested designs
Random slopes at the subject- and cluster-level.
Account for missing data/dropout.
Unbalanced designs (both unequal cluster sizes, and treatment groups).
Calculate the design effect, and estimated type I error when the third-level is ignored.
Fast analytical power calculations for all supported designs.
Explore bias, Type I error and model misspecification using. convenient simulation methods
Few clusters; accurate power analysis even with very few clusters, by using Satterthwaite's degrees of freedom approximation.
Create power curves to guide power analysis and help with optimal design of sample sizes at each level.

Type vignette("two-level", package = "powerlmm"), or vignette("three-level", package = "powerlmm") to see a tutorial on using powerlmm to calculate power. See all available vignettes by typing vignette(package = "powerlmm").

Kristoffer Magnusson

Maintainer: Kristoffer Magnusson <hello@kristoffer.email>

study_parameters, get_power

powerlmm documentation built on May 2, 2019, 3:10 a.m.