EasyStat

Automated Statistical Analysis, Visualization, and Multi-Format Narrative Reporting in R

Authors: Mr. Mahesh Divakaran & Dr. Gunjan Singh (Amity School of Applied Sciences, Amity University Lucknow) - Prof. Dr. Jayadevan Shreedharan (Gulf Medical University)

Overview

EasyStat bridges the gap between statistical output and actionable insight. A single function call delivers the statistical result, a plain-language narrative interpretation, and publication-ready tables, rendered in the RStudio Viewer, the R console, or Microsoft Word.

User-facing p-values are reported as percentages rounded to 4 decimal places, while raw model objects still retain the original numeric p-values for advanced use.

Installation

# From CRAN (when available)
install.packages("EasyStat")

# Development version from GitHub
# install.packages("devtools")
devtools::install_github("itsmdivakaran/Easystat")

# From local source
install.packages("path/to/EasyStat", repos = NULL, type = "source")

Quick Start

library(EasyStat)

# Linear regression with narrative
easy_regression(mpg ~ wt + hp, data = mtcars)

# Logistic regression with odds ratios
easy_logistic_regression(am ~ mpg + wt, data = mtcars)

# t-test
easy_ttest(mpg ~ am, data = mtcars)

# One-way ANOVA
easy_anova(Sepal.Length ~ Species, data = iris)

# Descriptive statistics for multiple variables
easy_describe(mtcars, vars = c("mpg", "hp", "wt"))

# Correlation heatmap
easy_correlation_heatmap(mtcars, vars = c("mpg", "hp", "wt", "qsec", "drat"))

# Export any result to Word
result <- easy_logistic_regression(am ~ mpg + wt, data = mtcars)
export_to_word(result, file = "report.docx", title = "Transmission Model",
               author = "Mahesh Divakaran, Gunjan Singh, Jayadevan Shreedharan")

Four-Step Pipeline

Step	Module	Role
1	Core Statistical Engine	Wraps `lm()`, `glm()`, `t.test()`, `aov()`, `chisq.test()`, `var.test()`, `cor.test()`
2	Metric Extractor	Uses model summaries and `broom` helpers to extract p-values, effect sizes, CIs, and fit metrics
3	Narrative Generator Module	Applies conditional logic to produce plain-language explanations
4	Unified Result Object	Returns `easystat_result` S3 objects with tables, narrative, and optional plots

Function Reference

Descriptive Statistics

Function	Description
`easy_describe()`	21-statistic summary for one or more numeric variables
`easy_group_summary()`	Stratified descriptives by a grouping factor

Regression Models

Function	Model	Key Output
`easy_regression()`	Linear regression	R-squared, ANOVA table, diagnostics, influential observations
`easy_logistic_regression()`	Binary logistic regression	Odds ratios, OR CIs, classification table, McFadden pseudo-R2

Inferential Tests

Function	Test	Effect Size
`easy_ttest()`	Independent / one-sample t-test	Cohen’s d
`easy_anova()`	One-way ANOVA with post-hoc context	eta-squared
`easy_chisq()`	Chi-square independence and GOF	Cramér’s V
`easy_ztest()`	One- and two-sample z-test	Cohen’s d
`easy_ftest()`	F-test for equality of variances	Variance ratio + CI
`easy_correlation()`	Pearson / Spearman / Kendall correlation and matrix	r, r-squared
`easy_wilcox()`	Wilcoxon rank-sum / signed-rank test	Median comparison + CI
`easy_kruskal()`	Kruskal-Wallis test	Rank-based eta-squared

Visualizations

Function	Plot type
`easy_histogram()`	Histogram with normal-curve overlay
`easy_boxplot()`	Grouped box-and-whisker plot
`easy_scatter()`	Scatter plot with regression line and R-squared
`easy_barplot()`	Count or mean (+/- SE) bar chart
`easy_qqplot()`	Q-Q normality plot
`easy_density()`	Kernel density curve, optionally grouped
`easy_correlation_heatmap()`	Annotated pairwise correlation heatmap
`easy_regression_diagnostics()`	Fitted-vs-residuals diagnostic plot
`easy_odds_ratio_plot()`	Logistic regression odds-ratio plot
`easy_autoplot()`	Smart dispatcher that picks the right plot for a result

Theme & Export

Function	Description
`theme_easystat()`	Consistent ggplot2 theme for all plots
`export_to_word()`	Formatted `.docx` report with flextable and officer

Output Modes

Mode	Trigger
RStudio HTML Viewer	Auto-detected in interactive sessions
Console	Scripts, terminals, non-interactive sessions
Word `.docx`	`export_to_word()`

Running the Smoke Test

source(system.file("smoke_test.R", package = "EasyStat"))

Citation

If you use EasyStat in your research, please cite:

Divakaran M., Singh G., & Shreedharan J. (2026). EasyStat: Automated Statistical Analysis, Visualization and Multi-Format Narrative Reporting in R (Version 2.0.0). Amity University Lucknow & Gulf Medical University. https://itsmdivakaran.github.io/Easystat/index.html