semnova

Latent repeated-measures ANOVA in R. semnova fits factorial within- and between-subject designs as structural equation models, so the dependent variable and any covariates can be latent rather than manifest composites.

The method is described in Langenberg et al. (2020), doi:10.1080/00273171.2020.1803038.

Installation

# install.packages("remotes")
# remotes::install_github("langenberg/semnova")
install.packages("semnova", repos = c('https://langenberg.r-universe.dev', 'https://cloud.r-project.org'))

Plots need a few more packages:

install.packages(c("ggplot2", "ggraph", "tidygraph", "magick"))

Usage

library(semnova)

fit <- semnova(
    data = reading_manifest,
    id = "id",
    dv = "re_fix_time",
    withins = c("grade", "sentence"),
    contrasts_arg = list(grade = "contr.sum", sentence = "contr.sum")
)

fit                     # model description and ANOVA table
summary(fit)            # plus the per-contrast estimates
plot(fit, terms = "grade")
dag(fit)                # path diagram

Add betweens for a multigroup model, covariates for covariates, and indicator to make the dependent variable latent.

Documentation

vignette("semnova")     # every feature, with worked examples
vignette("glh")         # the general linear hypothesis behind each test

License

GPL (>= 2).