esci: Estimation Statistics with Confidence Intervals esci logo

CRAN status Download count

esci provides student-friendly tools for estimation statistics:

esci is both an R package and a module in jamovi. If you’re looking for the R package, stay here. If you want esci in jamovi, download and install jamovi and then use the module library to add esci.

Leave comments, bug reports, suggestions, and questions about esci here

esci is still under development; expect breaking changes in the future especially for the visualization functions. If you need production-ready estimation, turn to statpsych

esci is built on top of statpsych and metafor. That is, almost all of the statistical calculations are passed off to these packages. The only exception is for confidence intervals for Cohen’s d (see documentation). Why does esci exist, then?

The visualizations produced by esci are exquisite in a large part because of the lovely ggdist package by Matthew Kay.

Installation

Assuming submission to CRAN goes well, you will be able to install esci with:

install.packages("esci")

Or, get the stable branch directly from github

# install.packages("devtools")
devtools::install_github('rcalinjageman/esci')

Or, try out the development branch:

# install.packages("devtools")
devtools::install_github('rcalinjageman/esci',  branch = "development")

Roadmap

Example

library(esci)

data("data_penlaptop1")
estimate <- estimate_mdiff_two(data_penlaptop1, transcription, condition)
plot_mdiff(estimate)

Example difference plot