AssumpSure

🎯 Purpose

AssumpSure is a user-friendly R Shiny application that helps researchers validate statistical assumptions and select appropriate tests before analysis, ensuring valid, transparent, and reproducible results.

Misused statistical tests due to overlooked or misinterpreted assumptions can undermine scientific findings.

With AssumpSure, you can:

• Validate statistical assumptions automatically.
• Choose the right test for your data with confidence.
• Generate publication-ready plots and summaries in one click, demonstrating transparent test selection to reviewers and readers.

---

👥 Who is it for?

AssumpSure is perfect for:

• Researchers with limited statistical expertise.
• Students, clinicians, and early-career scientists seeking a code-free analysis workflow.
• Teams needing reproducible, publication-ready statistical outputs.
• Anyone unsure if their data meets test assumptions.

---

📦 Key Features

• Automatically check test assumptions (normality, homogeneity, homoscedasticity, collinearity, etc.).
• Evaluate your test selection and recommend alternatives when appropriate.
• Apply Log, Yeo-Johnson, or Box-Cox transformations to non-normal data for LM and LMM, with before-and-after visualizations.
• Support a wide range of statistical methods.
• Generate boxplots, effect plots, and diagnostic plots for transparent reporting.
• Export plots and tables for reports or publications with one click.
• Provide an intuitive interface for users with no programming experience.

---

🧪 Supported Statistical Methods

• T-tests (independent & paired)
• Mann-Whitney U test
• Wilcoxon signed-rank test
• One-way ANOVA
• Kruskal-Wallis test
• Chi-square and Fisher’s exact tests
• Correlation analyses (Pearson, Spearman, Kendall, and Biweight midcorrelation)
• Linear and linear mixed-effects models
• Logistic, multinomial, Poisson and negative binomial regression

---

📁 Installation

Important: Please use R 4.2 or newer.

CRAN

install.packages("AssumpSure")

GitHub

# Install devtools if not already installed
install.packages("devtools")
library(devtools)

# Install AssumpSure package
# devtools::install_github("Ahmedbargheet/AssumpSure")
install.packages("AssumpSure", repos = c('https://ahmedbargheet.r-universe.dev', 'https://cloud.r-project.org'))

---

🚀 Launch the App

Either with

library(AssumpSure)
launch_app()

or

AssumpSure::launch_app()

---

💻 Use Online

You can use AssumpSure without installing it (performance is slower online):
https://ahmed-bargheet.shinyapps.io/AssumpSure/

---

📂 Data Format Guidelines

• Use long format: one row per subject/timepoint (see sample CSV included with the app)
• For longitudinal studies, name your time variable exactly timepoint

---

📖 Citation

If you use AssumpSure in your research, please cite it as below.

Bargheet, A. (2025). AssumpSure: a user-friendly R Shiny package for automated validation of statistical assumptions and appropriate test selection. Journal of Open Source Software, 10(115), 9286. https://doi.org/10.21105/joss.09286

---

📚 Reproducible Examples

The examples below demonstrate how to use each analysis module in AssumpSure. The infants.csv dataset is downloadable within the app (Help tab), while the remaining example datasets are available here on GitHub. Screenshots display the expected outputs for verifying functionality and reproducibility.

1. Independent t-test

Dataset: t-test.csv
Steps:

• Launch the app:

AssumpSure::launch_app()

• Upload t-test.csv in the Continuous Data Tests tab.
• Choose Independent T-test.
• Select the numeric value and Treatment as a categorical value, and click Check Assumptions.

2. One-way ANOVA

Dataset: plantgrowth.csv

• Upload plantgrowth.csv in the Continuous Data Tests tab.
• Choose One-way ANOVA.
• Choose the numeric value and the categorical value, and click Check Assumptions.
• Then click on Run Test, and choose p-value adjustment method or Tukey HSD, and click on Run the Selected Option
• Optionally, click Plot Boxplot to visualize group differences.

3. Fisher's exact test

Dataset: chi_fisher.csv

• Upload chi_fisher.csv in the Fisher & Chi-square tab.
• Choose sex as first categorical variable and location as second categorical variable.
• Choose Fisher's exact test from the Choose test drop menu and click on Run Test.
• Choose Benjamini-Hochberg
• Optionally, click Plot to visualize groups.

3. Linear Model

Dataset: infants.csv

• Upload infants.csv in the Regression Models tab.
• Choose weight as dependent variable and Treatment, gender, country, nutrients, and diarrhea as independent variables.
• Choose Inverse normal from the Transform dependent variable drop menu and click on Check Assumptions.
• Then click on run LM and Plot Forest.

4. Correlation

Dataset: correlation.csv

• Upload correlation.csv in the Correlation tab.
• Choose Pearson from the Correlation Method drop menu and click Check Assumptions.

5. Correlation for compositional data

Dataset: bacteria_for_correlation.csv

• Upload bacteria_for_correlation.csv in the Correlation tab.
• Tick Apply CLR transformation, and set Remove features to 20% using the slider
• Choose Spearman from the Correlation Method drop menu and click Run Correlation.
• Select at least five features to display the heatmap.
• The user can download the heatmap, matrix, and table using the download buttons.

---

📚 Need Help?

For feedback or questions, feel free to contact me at ahmed.bargheet@yahoo.com