Durga: an R package for effect size estimation and visualization

Area of Science:

• Statistics
• Data Visualization
• Science Communication

Background:

• Overreliance on p-values in statistical analysis hinders effective science communication.
• Effect sizes are crucial for quantitative research clarity but often underutilized.
• Current reporting of effect sizes is mainly limited to tables, lacking visual impact.

Purpose of the Study:

• Introduce the Durga R package for estimating and plotting effect sizes.
• Provide a user-friendly tool for both paired and unpaired group comparisons.
• Facilitate more effective data analysis and science communication.

Main Methods:

• Developed the Durga R package for statistical analysis.
• Implemented functions to estimate unstandardized and standardized effect sizes.
• Integrated bootstrapped confidence intervals for effect size estimation.
• Combined effect size visualizations with traditional plotting methods.

Main Results:

• Durga facilitates flexible estimation of effect sizes for various statistical methods.
• The package offers extensive options for aesthetic and informative effect size plotting.
• Demonstrated a workflow for estimating and plotting effect sizes using example datasets.

Conclusions:

• Durga R package offers a powerful and accessible solution for effect size estimation and visualization.
• Promotes enhanced clarity and effectiveness in scientific data analysis and communication.
• Encourages a shift from p-value dependency towards robust effect size reporting.