Installation

First, install R and R studio. Then, copy and paste the following lines in the console:

install.packages("remotes")
remotes::install_github("easystats/report") # You only need to do that once
library("report") # Load the package every time you start R

Great! The report package is now installed and loaded in your session.

Supported Objects

The report package works in a two step fashion: - First, you create a report object with the report() function. - Second, this report object can be displayed either textually (the default output) or as a table, using as.data.frame(). Moreover, you can also access a more compact version of the report using summary() on the report object.

Dataframes

If an entire dataframe is supplied, report will provide descriptive statistics for all columns:

report(iris)
# The data contains 150 observations of the following 5 variables:
#   - Sepal.Length: n = 150, Mean = 5.84, SD = 0.83, Median = 5.80, MAD = 1.04, range: [4.30, 7.90], Skewness = 0.31, Kurtosis = -0.55, 0% missing
#   - Sepal.Width: n = 150, Mean = 3.06, SD = 0.44, Median = 3.00, MAD = 0.44, range: [2, 4.40], Skewness = 0.32, Kurtosis = 0.23, 0% missing
#   - Petal.Length: n = 150, Mean = 3.76, SD = 1.77, Median = 4.35, MAD = 1.85, range: [1, 6.90], Skewness = -0.27, Kurtosis = -1.40, 0% missing
#   - Petal.Width: n = 150, Mean = 1.20, SD = 0.76, Median = 1.30, MAD = 1.04, range: [0.10, 2.50], Skewness = -0.10, Kurtosis = -1.34, 0% missing
#   - Species: 3 levels, namely setosa (n = 50, 33.33%), versicolor (n = 50, 33.33%) and virginica (n = 50, 33.33%)

Grouped Dataframes

The dataframe can also be a grouped dataframe (from dplyr), in which case report would return a separate report for each level of the grouping variable. Additionally, instead of textual summary, report also allows one to return a tabular summary using the report_table() function:

library(dplyr)

iris %>%
  group_by(Species) %>%
  report_table()
# Group      |     Variable | n_Obs | Mean |   SD | Median |  MAD |  Min |  Max | Skewness | Kurtosis | n_Missing
# ---------------------------------------------------------------------------------------------------------------
# versicolor | Sepal.Length |    50 | 5.94 | 0.52 |   5.90 | 0.52 | 4.90 | 7.00 |     0.11 |    -0.53 |         0
# versicolor |  Sepal.Width |    50 | 2.77 | 0.31 |   2.80 | 0.30 | 2.00 | 3.40 |    -0.36 |    -0.37 |         0
# versicolor | Petal.Length |    50 | 4.26 | 0.47 |   4.35 | 0.52 | 3.00 | 5.10 |    -0.61 |     0.05 |         0
# versicolor |  Petal.Width |    50 | 1.33 | 0.20 |   1.30 | 0.22 | 1.00 | 1.80 |    -0.03 |    -0.41 |         0
# virginica  | Sepal.Length |    50 | 6.59 | 0.64 |   6.50 | 0.59 | 4.90 | 7.90 |     0.12 |     0.03 |         0
# virginica  |  Sepal.Width |    50 | 2.97 | 0.32 |   3.00 | 0.30 | 2.20 | 3.80 |     0.37 |     0.71 |         0
# virginica  | Petal.Length |    50 | 5.55 | 0.55 |   5.55 | 0.67 | 4.50 | 6.90 |     0.55 |    -0.15 |         0
# virginica  |  Petal.Width |    50 | 2.03 | 0.27 |   2.00 | 0.30 | 1.40 | 2.50 |    -0.13 |    -0.60 |         0

Correlations, t-test, and Wilcox test

report can also be used to provide automated summaries for statistical model objects from correlation, t-tests, etc.

report(t.test(formula = wt ~ am, data = mtcars))
# Effect sizes were labelled following Cohen's (1988) recommendations.
# 
# The Welch Two Sample t-test testing the difference of wt by am (mean in group 0 = 3.77, mean in group 1 = 2.41) suggests that the effect is positive, statistically significant, and large (difference = 1.36, 95% CI [0.85, 1.86], t(29.23) = 5.49, p < .001; Cohen's d = 2.03, 95% CI [1.13, 2.91])
x <- c(1.83,  0.50,  1.62,  2.48, 1.68, 1.88, 1.55, 3.06, 1.30)
y <- c(0.878, 0.647, 0.598, 2.05, 1.06, 1.29, 1.06, 3.14, 1.29)

report(wilcox.test(x, y, paired = TRUE))

Regression models

Linear regression (lm)

We will start out simple: a simple linear regression

model <- lm(wt ~ am + mpg, data = mtcars)

report(model)
# We fitted a linear model (estimated using OLS) to predict wt with am and mpg (formula: wt ~ am + mpg). The model explains a statistically significant and substantial proportion of variance (R2 = 0.80, F(2, 29) = 57.66, p < .001, adj. R2 = 0.79). The model's intercept, corresponding to am = 0 and mpg = 0, is at 5.74 (95% CI [5.11, 6.36], t(29) = 18.64, p < .001). Within this model:
# 
#   - The effect of am is statistically significant and negative (beta = -0.53, 95% CI [-0.94, -0.11], t(29) = -2.58, p = 0.015; Std. beta = -0.27, 95% CI [-0.48, -0.06])
#   - The effect of mpg is statistically significant and negative (beta = -0.11, 95% CI [-0.15, -0.08], t(29) = -6.79, p < .001; Std. beta = -0.71, 95% CI [-0.92, -0.49])
# 
# Standardized parameters were obtained by fitting the model on a standardized version of the dataset.

anova (aov)

And its close cousin ANOVA is also covered by report:

model <- aov(wt ~ am + mpg, data = mtcars)

report(model)
# The ANOVA (formula: wt ~ am + mpg) suggests that:
# 
#   - The main effect of am is statistically significant and large (F(1, 29) = 69.21, p < .001; Eta2 (partial) = 0.70, 95% CI [0.54, 1.00])
#   - The main effect of mpg is statistically significant and large (F(1, 29) = 46.12, p < .001; Eta2 (partial) = 0.61, 95% CI [0.42, 1.00])
# 
# Effect sizes were labelled following Field's (2013) recommendations.

General Linear Models (GLMs) (glm)

model <- glm(vs ~ mpg + cyl, data = mtcars, family = "binomial")

report(model)
# We fitted a logistic model (estimated using ML) to predict vs with mpg and cyl (formula: vs ~ mpg + cyl). The model's explanatory power is substantial (Tjur's R2 = 0.67). The model's intercept, corresponding to mpg = 0 and cyl = 0, is at 15.97 (95% CI [-2.71, 44.69], p = 0.147). Within this model:
# 
#   - The effect of mpg is statistically non-significant and negative (beta = -0.16, 95% CI [-0.71, 0.34], p = 0.496; Std. beta = -0.98, 95% CI [-4.28, 2.03])
#   - The effect of cyl is statistically significant and negative (beta = -2.15, 95% CI [-5.19, -0.54], p = 0.047; Std. beta = -3.84, 95% CI [-9.26, -0.97])
# 
# Standardized parameters were obtained by fitting the model on a standardized version of the dataset. 95% Confidence Intervals (CIs) and p-values were computed using

Linear Mixed-Effects Models (merMod)

library(lme4)

model <- lmer(Reaction ~ Days + (Days | Subject), data = sleepstudy)

report(model)
# We fitted a linear mixed model (estimated using REML and nloptwrap optimizer) to predict Reaction with Days (formula: Reaction ~ Days). The model included Days and Subject as random effects (formula: ~Days | Subject). The model's total explanatory power is substantial (conditional R2 = 0.80) and the part related to the fixed effects alone (marginal R2) is of 0.28. The model's intercept, corresponding to Days = 0, is at 251.41 (95% CI [237.94, 264.87], t(174) = 36.84, p < .001). Within this model:
# 
#   - The effect of Days is statistically significant and positive (beta = 10.47, 95% CI [7.42, 13.52], t(174) = 6.77, p < .001; Std. beta = 0.54, 95% CI [0.38, 0.69])
# 
# Standardized parameters were obtained by fitting the model on a standardized version of the dataset. 95% Confidence Intervals (CIs) and p-values were computed using