Skip to contents

statsreportr

Source: vignettes/statsreportr.Rmd
statsreportr.Rmd

The goal of statsreportr is to make it easy to report the results of statistical analyses in inline code of a R Markdown or Quarto document. It is designed to work with the rstatix package, but can also use some other inputs, and when rendering to a pdf. The envisioned usecase is to use rstatix to perform the statistical tests, and then use statsreportr to format the results in a way that is easy to include in a scientific or statistical report.

Installation

You can install the development version of statsreportr from GitHub with:

# install.packages("pak")
pak::pak("KyleOfCanada/statsreportr")

Examples

Here are examples of the functions in statsreportr. In these examples they are outputing to the R console, but they are intended to be used inline in a R Markdown or Quarto document.

library(statsreportr)

# ANOVAs
results <- mtcars |>
  rstatix::anova_test(mpg ~ cyl)

format_p(results$p)
#> [1] "< 0.0001"

results2 <- mtcars |>
  rstatix::anova_test(mpg ~ cyl * carb)

results3 <- aov(mpg ~ cyl * carb, data = mtcars)

report_anova(results2)
#> [1] "*F*~(1,28)~ = 47.2, *p* < 0.0001, $\\eta^2_G$ = 0.628"
report_anova(results2, "carb")
#> [1] "*F*~(1,28)~ = 1.55, *p* = 0.223, $\\eta^2_G$ = 0.052"
report_anova(results2, 2)
#> [1] "*F*~(1,28)~ = 1.55, *p* = 0.223, $\\eta^2_G$ = 0.052"
report_anova(results3, 2)
#> [1] "*F*~(1,28)~ = 1.55, *p* = 0.223, $\\eta^2_G$ = 0.052"

# t-tests
results3 <- mtcars |>
  rstatix::t_test(mpg ~ am)

report_t(results3)
#> [1] "*t*~(18.3)~ = -3.77, *p* = 0.00137"

results4 <- mtcars |>
  rstatix::t_test(mpg ~ cyl)

cohensd4 <- mtcars |>
  rstatix::cohens_d(mpg ~ cyl)

report_t(results4,
  effect = c("6", "8"),
  cohensd = cohensd4, cohens_magnitude = TRUE
)
#> [1] "*t*~(18.5)~ = 5.29, adjusted *p* < 0.0001, *d* = 2.23, indicating a large effect"

# Descriptive stats

mtcars |> report_mean_sd(mpg, cyl, effect = "6")
#> [1] "19.7 ± 1.45"

mtcars |> report_mean_sem(mpg, cyl, effect = "6")
#> [1] "19.7 ± 0.549"

When used with inline code in a R Markdown or Quarto document, the following inline code:

The results of the t test showed a significant difference in the mpg of 6 (`r mtcars |> report_mean_sd(mpg, cyl, effect = "6")`) vs 8 (`r mtcars |> report_mean_sd(mpg, cyl, effect = "8")`) cylinder cars (`r report_t(results4, effect = c("6", "8"), cohensd = cohensd4, cohens_magnitude = TRUE)`).

Will render as:

The results of the t test showed a significant difference in the mpg of 6 (19.7 ± 1.45) vs 8 (15.1 ± 2.56) cylinder cars (t(18.5) = 5.29, adjusted p < 0.0001, d = 2.23, indicating a large effect).

Functions

The following functions are available in the package: