jjstatsplot

jjstatsplot is an open-source R package and jamovi module developed by Serdar Balci that wraps the ggstatsplot framework (created by Indrajeet Patil) into an intuitive, point-and-click graphical interface for jamovi and a high-level programmatic interface for R.

Traditional scientific workflows typically decouple statistical hypothesis testing from data visualization, forcing researchers to manually run statistical tests in one step and subsequently annotate plots or write captions in another. jjstatsplot combines statistical analysis and data visualization into single, cohesive, publication-ready graphics that automatically embed test statistics, $p$-values, effect sizes with confidence intervals, sample sizes, and Bayesian factors directly in plot annotations.

• GitHub Repository: sbalci/jjstatsplot

• Documentation & Vignettes: serdarbalci.com/jjstatsplot

• ggstatsplot Upstream Guide: indrapatil.com/ggstatsplot

• Bug Reports & Issues: ClinicoPathJamoviModule/issues

• ClinicoPath Suite Citation: 10.5281/zenodo.3997188

• License: GPL (>= 2)

---

Key Capabilities & Architecture

1. Integrated Statistical Hypotheses: Every supported visualization offers automatic calculation and subtitle reporting across four major statistical frameworks:

   • Parametric: Student's $t$-test, Welch's $t$-test, Fisher's one-way ANOVA, Pearson's product-moment correlation ($r$).

   • Non-parametric: Mann-Whitney $U$ test, Kruskal-Wallis one-way ANOVA, Wilcoxon signed-rank test, Friedman test, Spearman's rank correlation ($\rho$).

   • Robust: Trimmed-means comparisons (Yuen's test), Huber and percentage-bend robust correlation coefficients, and robust heteroscedastic ANOVAs via WRS2.

   • Bayesian: Bayesian $t$-tests, one-way ANOVA, contingency table tests, and correlation analyses via BayesFactor and statsExpressions, reporting Bayes Factors ($\log_e(BF_{10})$ or $BF_{10}$) and posterior estimates.

2. Dual Execution Interfaces:

   • jamovi Graphical Interface: Integrated natively into the jamovi Analyses ribbon, allowing clinicians and researchers to drag and drop variables, toggle between statistical paradigms, customize aesthetics, and export vector graphics without writing code.

   • R Programmatic Interface: Exportable, clean functions (jjhistostats(), jjbetweenstats(), etc.) usable in scripts, R Markdown, and Quarto reports.

3. Multi-Variable & Grouped Analysis:

   • Dual-Mode Operation: Functions process either single dependent variables or multiple variables simultaneously.

   • Automatic Grouping/Faceting: Facets plots seamlessly across categorical covariates (e.g., patient subgroups, histological subtypes, or treatment arms).

4. Rich Aesthetic & Theme Customization:

   • Pre-packaged color palettes from RColorBrewer, viridis, and ggprism.

   • Configurable point jittering, violin/box layer ordering, and outlier labeling.

   • Dynamic canvas resizing to accommodate complex labels, long axis titles, and detailed statistical subtitles.

---

The 18 Analysis Types

Category	Function / Analysis	Description	Statistical Details Embedded
Continuous Distributions	Histogram (jjhistostats)	Distribution visualization with density curves	Shapiro-Wilk test, robust central tendencies, parametric/Bayesian tests
Continuous vs Continuous	Scatter Plot (jjscatterstats)	Bivariate relationship with marginal distributions	Correlation coefficients ($r$, $\rho$), linear/robust regression, $p$-values, $BF_{10}$
Continuous vs Continuous	Correlation Matrix (jjcorrmat)	Matrix of pairwise correlations across multiple variables	Significance levels, correlation coefficients, $p$-value matrices
Continuous vs Continuous	Hull Plot (hullplot)	2D scatter visualization with convex/concave hulls	Convex hull polygonal boundaries, cluster and group membership
Categorical vs Continuous	Box-Violin Plot (jjbetweenstats)	Between-groups comparison across independent cohorts	Student's/Welch's ANOVA, Kruskal-Wallis, Yuen's trimmed test, post-hoc tests
Categorical vs Continuous	Box-Violin Plot (jjwithinstats)	Within-subjects / repeated measures comparison	Repeated measures ANOVA, Friedman test, paired post-hoc tests
Categorical vs Continuous	Dot Chart (jjdotplotstats)	Mean and median comparisons across categories	Confidence intervals, effect sizes, point estimates
Categorical vs Categorical	Bar Chart (jjbarstats)	Two-way cross-tabulation and composition	Pearson's $\chi^2$, Fisher's exact test, Cramér's $V$, contingency coefficients
Categorical vs Categorical	Pie Chart (jjpiestats)	Categorical composition and proportions	Goodness-of-fit $\chi^2$, proportions test, effect sizes
Distribution Parts-to-Whole	Waffle Chart (jjwaffle)	Square grid / waffle visualization of counts/percentages	Proportional breakdown across discrete classes
Advanced Distributions	Raincloud Plot (raincloud)	Hybrid raw jittered points + boxplot + split violin density	Kernel density, quartiles, individual datapoint distributions
Advanced Distributions	Advanced Raincloud (advancedraincloud)	Enhanced raincloud plots with longitudinal tracking	Repeated measures paths, multi-group stratifications
Distribution Comparison	Ridge Plots (jjridges)	Staggered ridgeline density curves via ggridges	Overlapping group densities, quantile shading, multimodal checks
Network & Flow	Arc Diagrams (jjarcdiagram)	Node-link relationships with curved arcs	Connection weights, node rankings, directional connections
Segmented Proportions	Segmented Bar (jjsegmentedtotalbar)	Stacked bar with segment-level and overall counts	Subgroup proportions paired with total sample counts
Time Series / Trends	Line Chart (linechart)	Trajectories and continuous progressions	Rate of change, mean trends, confidence ribbons
Ranked Data	Lollipop Chart (lollipop)	Clean stem-and-dot comparison of discrete categories	Ranked differences, divergence from baseline or grand mean
Extended Multi-Method	Stats Plot 2 (statsplot2)	Composite multi-panel statistical displays	Harmonized multi-variable statistical comparisons

---

Quick Start in R

# Install from GitHub
if (!requireNamespace("devtools", quietly = TRUE)) install.packages("devtools")
devtools::install_github("sbalci/jjstatsplot")

library(jjstatsplot)

# 1. Independent group comparison with embedded statistics
jjbetweenstats(
  data = iris,
  dep = "Sepal.Length",
  group = "Species",
  type = "parametric",     # "parametric", "nonparametric", "robust", or "bayes"
  pairwise.comparisons = TRUE,
  p.adjust.method = "bonferroni"
)

# 2. Correlation scatter plot with marginal distributions
jjscatterstats(
  data = mtcars,
  dep = "wt",
  group = "mpg",
  type = "robust",
  marginal = TRUE
)

# 3. Categorical distribution and association test
jjbarstats(
  data = as.data.frame(Titanic),
  dep = "Survived",
  group = "Class",
  counts = "Freq"
)

---

Upstream Citations & Acknowledgments

When utilizing jjstatsplot in academic research, cite the module and its foundational packages:

• jjstatsplot / ClinicoPath Suite:

  Balci S. ClinicoPath jamovi Module. DOI: 10.5281/zenodo.3997188.

• ggstatsplot Framework:

  Patil I. (2021). Visualizations with statistical details: The 'ggstatsplot' approach. Journal of Open Source Software, 6(61), 3167. DOI: 10.21105/joss.03167.

• jamovi Platform:

  The jamovi project. jamovi (Computer Software). https://www.jamovi.org.