General 01: jamovi User Guide for Statistical Analysis

jamovi User Guide: A Statistics Crash Course for Pathologists and Clinicians

This comprehensive guide introduces jamovi as a powerful, user-friendly statistical software for medical research. jamovi provides point-and-click access to sophisticated statistical analyses without requiring programming knowledge, making it ideal for pathologists and clinicians who need to analyze research data.

Installation

Installing jjstatsplot in jamovi

1. Open jamovi
2. Access Modules: Click the “Modules” button (⊞) in the top menu
3. Open jamovi library: Select “jamovi library” from the dropdown
4. Search: Type “jjstatsplot” in the search box
5. Install: Click the “Install” button next to jjstatsplot
6. Restart: Close and reopen jamovi to activate the module

Verification

After installation, you should see “JJStatsPlot” in the main menu bar with submenus for different analysis types.

Understanding the Interface

Menu Organization

The jjstatsplot analyses are organized by data type:

• Continuous: Histogram
• Continuous vs Continuous: Scatter Plot, Correlation Matrix
  
• Categorical vs Continuous: Box-Violin Plots, Dot Charts
• Categorical vs Categorical: Bar Charts, Pie Charts
• Distribution: Waffle Charts

Common Interface Elements

All jjstatsplot analyses share similar interface components:

Variables Panel

• Dependent Variable(s): The main variable(s) you want to analyze
• Grouping Variable: Optional variable to split analysis by groups
• Grouping Variable for Plots: Creates separate plots for each group level

Options Panel

• Statistical Options: Choose test types, confidence levels
• Plot Options: Customize appearance, themes, colors
• Advanced Options: Fine-tune statistical parameters

Data Types and Requirements

Variable Types

jjstatsplot automatically detects variable types, but understanding them helps:

• Continuous: Numeric variables (age, height, score)
• Nominal: Categories without order (gender, color, treatment)
• Ordinal: Categories with order (education level, satisfaction rating)

Data Format Requirements

• Long format: Each row represents one observation
• Complete cases: Missing values are handled automatically
• Appropriate sample sizes: Minimum recommendations vary by analysis

Analysis Walkthrough

1. Histogram Analysis

Use case: Explore the distribution of a continuous variable

Steps: 1. Navigate to JJStatsPlot → Continuous → Histogram 2. Move your continuous variable to Dependent Variable 3. Optional: Add a Grouping Variable for separate histograms 4. Customize in Options: - Statistical: Choose normality test type - Plot: Adjust bins, colors, theme - Labels: Add title, axis labels

Example Output: - Histogram with density curve - Normality test results - Descriptive statistics overlay

2. Scatter Plot Analysis

Use case: Examine relationship between two continuous variables

Steps: 1. Go to JJStatsPlot → Continuous vs Continuous → Scatter Plot 2. Set Dependent Variable (Y-axis) 3. Set Grouping Variable (X-axis) 4. Optional: Add Grouping Variable for Plots for separate panels 5. Configure Options: - Statistical: Correlation test, regression line - Plot: Point style, smoothing method - Marginal plots: Add distribution plots to margins

Example Output: - Scatter plot with regression line - Correlation coefficient and significance - Confidence intervals - Marginal distribution plots (optional)

3. Box-Violin Plots (Between Groups)

Use case: Compare continuous variable across different groups

Steps: 1. Select JJStatsPlot → Categorical vs Continuous → Box-Violin Plots (Between Groups) 2. Add continuous variable to Dependent Variable 3. Add grouping variable to Grouping Variable 4. Adjust Options: - Statistical: Choose comparison test (t-test, ANOVA, non-parametric) - Plot: Combine box and violin plots - Pairwise: Enable post-hoc comparisons

Example Output: - Combined box and violin plots - Statistical test results - Effect size measures - Pairwise comparison results

4. Correlation Matrix

Use case: Explore relationships among multiple continuous variables

Steps: 1. Navigate to JJStatsPlot → Continuous vs Continuous → Correlation Matrix 2. Select multiple variables for Dependent Variables 3. Optional: Add Grouping Variable for Plots 4. Customize Options: - Statistical: Correlation method (Pearson, Spearman) - Plot: Color scheme, significance marking - Matrix type: Full, upper, or lower triangle

Example Output: - Color-coded correlation matrix - Significance indicators - Correlation coefficients

5. Bar Charts

Use case: Visualize frequency or proportions of categorical variables

Steps: 1. Go to JJStatsPlot → Categorical vs Categorical → Bar Charts 2. Set Dependent Variable (categories to count) 3. Optional: Set Grouping Variable for grouped bars 4. Configure Options: - Statistical: Chi-square test, effect size - Plot: Bar orientation, colors - Labels: Show counts, percentages

Example Output: - Bar chart with counts/proportions - Chi-square test results - Effect size (Cramér’s V)

Advanced Features

Grouped Analysis

Most analyses support grouped analysis:

1. Multiple Dependent Variables: Creates subplot for each variable
2. Grouping Variable for Plots: Creates separate plot for each group level
3. Combination: Multiple variables × multiple groups = grid of plots

Theme Options

Choose between: - jamovi theme: Clean, publication-ready appearance - ggstatsplot theme: Rich statistical annotations and colors

Export Options

1. Copy Plot: Right-click plot → Copy
2. Save: File → Export → choose format (PNG, PDF, SVG)
3. Results: Copy statistical output from results panel

Troubleshooting Common Issues

“No data to plot” Error

• Check: Ensure variables are selected correctly
• Missing data: Variables with all missing values can’t be plotted
• Variable type: Ensure variable types match analysis requirements

Statistical Test Failures

• Sample size: Some tests require minimum sample sizes
• Assumptions: Check if data meets test assumptions
• Alternative tests: Try non-parametric alternatives

Plot Display Issues

• Refresh: Try refreshing the analysis
• Simplify: Remove grouping variables temporarily
• Update: Ensure jamovi and module are up to date

Tips for Better Visualizations

1. Variable Selection

• Meaningful groupings: Choose grouping variables with 2-8 levels
• Sufficient data: Ensure adequate observations per group
• Relevant comparisons: Select variables that make theoretical sense

2. Customization

• Titles: Add descriptive titles and axis labels
• Colors: Use color schemes appropriate for your audience
• Theme: Choose theme based on publication requirements

3. Statistical Interpretation

• Effect sizes: Always interpret alongside p-values
• Assumptions: Check and report assumption violations
• Multiple comparisons: Consider correction when doing many tests

Example Workflow

Research Question: “Does engine type affect fuel efficiency?”

1. Data: Load dataset with mpg (continuous) and vs (engine type, categorical)

2. Exploratory Analysis:

   • Histogram: Examine mpg distribution
   • Box-Violin Plot: Compare mpg between vs groups

3. Detailed Analysis:

   • Use Box-Violin Plots (Between Groups)
   • Set mpg as Dependent Variable
   • Set vs as Grouping Variable
   • Enable statistical tests and effect sizes

4. Interpretation:

   • Examine group differences in plots
   • Report statistical test results
   • Include effect size measures

Best Practices

Statistical Reporting

• Always report effect sizes alongside p-values
• Check and report assumption violations
• Use appropriate tests for your data type and distribution

Visualization Guidelines

• Keep plots simple and interpretable
• Use consistent color schemes across related analyses
• Include appropriate titles and labels
• Consider your audience (academic, clinical, general public)

Reproducibility

• Document your analysis choices
• Save jamovi files with descriptive names
• Export high-quality images for publications

Getting Help

Resources

• ggstatsplot documentation: indrajeetpatil.github.io/ggstatsplot
• jamovi community: forum.jamovi.org
• GitHub issues: github.com/sbalci/ClinicoPathJamoviModule/issues

Support

• Check jamovi forum for similar questions
• Report bugs via GitHub issues
• Include sample data and screenshots when asking for help

This guide provides a foundation for using jjstatsplot effectively in jamovi. Each analysis type offers extensive customization options - experiment with different settings to find what works best for your research needs.