Complete Analysis Gallery - All jjstatsplot Functions

ClinicoPath Development Team

2025-06-30

Complete jjstatsplot Analysis Gallery

This comprehensive guide demonstrates every analysis type available in jjstatsplot with practical examples, use cases, and statistical interpretations.

library(ClinicoPath)
library(dplyr)

# Load datasets
data(mtcars)
data(iris)

# Prepare example datasets
mtcars_clean <- mtcars %>%
  mutate(
    cyl = factor(cyl, labels = c("4-cyl", "6-cyl", "8-cyl")),
    am = factor(am, labels = c("Automatic", "Manual")),
    vs = factor(vs, labels = c("V-shaped", "Straight"))
  )

iris_clean <- iris %>%
  mutate(Species = factor(Species))

1. Histogram Analysis - jjhistostats()

Purpose

Explore the distribution of continuous variables with automatic normality testing and descriptive statistics.

When to Use

• Examine variable distributions before analysis
• Check normality assumptions
• Identify outliers and skewness
• Compare distributions across groups

Basic Usage

# Single variable distribution
hist_basic <- jjhistostats(
  data = mtcars_clean,
  dep = "mpg",
  grvar = NULL  # No grouping variable
)

hist_basic$plot

Advanced: Multiple Variables

# Multiple variables
hist_multi <- jjhistostats(
  data = iris,
  dep = c("Sepal.Length", "Sepal.Width", "Petal.Length"),
  grvar = NULL  # No grouping variable
)

hist_multi$plot

Grouped Analysis

# Separate histograms by group
hist_grouped <- jjhistostats(
  data = mtcars_clean,
  dep = "mpg",
  grvar = "cyl"
)

hist_grouped$plot2

Statistical Output

• Shapiro-Wilk normality test
• Descriptive statistics (mean, median, SD)
• Sample size information

Interpretation Tips

• Normal distribution: p > 0.05 in normality test
• Skewness: Asymmetric distribution patterns
• Outliers: Points far from main distribution

---

2. Scatter Plots - jjscatterstats()

Purpose

Examine relationships between two continuous variables with correlation analysis and regression fitting.

When to Use

• Explore bivariate relationships
• Test correlation hypotheses
• Visualize regression relationships
• Identify influential points

Basic Usage

# Basic scatter plot with correlation
scatter_basic <- jjscatterstats(
  data = mtcars_clean,
  dep = "mpg",
  group = "hp",
  grvar = NULL  # No grouping variable
)

scatter_basic$plot

Grouped Analysis

# Separate scatter plots by group
scatter_grouped <- jjscatterstats(
  data = mtcars_clean,
  dep = "mpg",
  group = "hp", 
  grvar = "cyl"
)

scatter_grouped$plot2

Statistical Output

• Pearson/Spearman correlation coefficient
• Confidence intervals
• Regression equation
• R-squared value

Interpretation Tips

• Strong correlation: |r| > 0.7
• Moderate correlation: 0.3 < |r| < 0.7
• Weak correlation: |r| < 0.3
• Linear relationship: Points follow straight line pattern

---

3. Box-Violin Plots (Between Groups) - jjbetweenstats()

Purpose

Compare continuous variables across independent groups with appropriate statistical tests.

When to Use

• Compare means/medians between groups
• Test group differences
• Visualize data distribution by group
• Assess homogeneity of variance

Basic Usage

# Compare groups
between_basic <- jjbetweenstats(
  data = iris_clean,
  dep = "Sepal.Length",
  group = "Species"
)

between_basic$plot

Multiple Dependent Variables

# Multiple variables comparison
between_multi <- jjbetweenstats(
  data = iris_clean,
  dep = c("Sepal.Length", "Petal.Length"),
  group = "Species"
)

between_multi$plot

Statistical Output

• ANOVA or Kruskal-Wallis test
• Post-hoc pairwise comparisons
• Effect size measures (eta-squared, Cohen’s d)
• Homogeneity of variance tests

Interpretation Tips

• Significant difference: p < 0.05
• Large effect: η² > 0.14 or d > 0.8
• Medium effect: η² > 0.06 or d > 0.5
• Small effect: η² > 0.01 or d > 0.2

---

4. Correlation Matrix - jjcorrmat()

Purpose

Visualize correlations among multiple continuous variables simultaneously.

When to Use

• Explore multivariate relationships
• Identify redundant variables
• Screen variables for analysis
• Data reduction decisions

Basic Usage

# Correlation matrix
corrmat_basic <- jjcorrmat(
  data = mtcars,
  dep = c("mpg", "hp", "wt", "qsec", "disp"),
  grvar = NULL  # No grouping variable
)

corrmat_basic$plot

Grouped Analysis

# Separate correlation matrices by group
corrmat_grouped <- jjcorrmat(
  data = iris,
  dep = c("Sepal.Length", "Sepal.Width", "Petal.Length", "Petal.Width"),
  grvar = "Species"
)

corrmat_grouped$plot2

Statistical Output

• Correlation coefficients matrix
• Significance testing for each correlation
• Confidence intervals

Interpretation Tips

• High correlation: |r| > 0.8 (consider multicollinearity)
• Pattern recognition: Look for clusters of correlated variables
• Significance: Starred values indicate p < 0.05

---

5. Dot Charts - jjdotplotstats()

Purpose

Display group means with confidence intervals and statistical comparisons.

When to Use

• Present group means clearly
• Show uncertainty (confidence intervals)
• Compare multiple groups
• Publication-ready group comparisons

Basic Usage

# Dot chart with group means
dot_basic <- jjdotplotstats(
  data = mtcars_clean,
  dep = "mpg",
  group = "cyl",
  grvar = NULL  # No grouping variable
)

dot_basic$plot

Multiple Variables

# Multiple dependent variables
dot_multi <- jjdotplotstats(
  data = iris_clean,
  dep = c("Sepal.Length"),
  group = "Species",
  grvar = NULL  # No grouping variable
)

dot_multi$plot

Statistical Output

• Group means and standard errors
• Confidence intervals
• Statistical test results
• Effect sizes

Interpretation Tips

• Non-overlapping CIs: Likely significant difference
• Large separation: Meaningful group differences
• Consistent patterns: Across multiple variables

---

6. Bar Charts - jjbarstats()

Purpose

Analyze categorical variables and their associations with frequency tests.

When to Use

• Display frequency distributions
• Test independence between categorical variables
• Visualize contingency tables
• Report categorical data summaries

Basic Usage

# Frequency bar chart
bar_basic <- jjbarstats(
  data = mtcars_clean,
  dep = "cyl",
  group = NULL,  # No grouping variable
  grvar = NULL  # No grouping variable
)

bar_basic$plot

Two-Way Analysis

# Two categorical variables
bar_twoway <- jjbarstats(
  data = mtcars_clean,
  dep = "cyl",
  group = "am",
  grvar = NULL  # No additional grouping variable
)

bar_twoway$plot

Statistical Output

• Chi-square test of independence
• Cramér’s V (effect size)
• Expected vs observed frequencies
• Standardized residuals

Interpretation Tips

• Significant association: p < 0.05 in chi-square test
• Strong association: Cramér’s V > 0.5
• Moderate association: Cramér’s V > 0.3
• Weak association: Cramér’s V > 0.1

---

7. Pie Charts - jjpiestats()

Purpose

Display proportional data with statistical testing for categorical variables.

When to Use

• Show proportions of categories
• Test goodness-of-fit
• Display survey results
• Present percentage distributions

Basic Usage

# Basic pie chart
pie_basic <- jjpiestats(
  data = mtcars_clean,
  dep = "cyl",
  group = NULL,  # No grouping variable
  grvar = NULL  # No grouping variable
)

pie_basic$plot1

Grouped Analysis

# Separate pie charts by group
pie_grouped <- jjpiestats(
  data = mtcars_clean,
  dep = "cyl",
  group = NULL,  # No grouping variable
  grvar = "am"
)

pie_grouped$plot2

Statistical Output

• Goodness-of-fit tests
• Frequency and percentage tables
• Chi-square statistics

Interpretation Tips

• Equal proportions: Uniform distribution
• Dominant category: One slice much larger
• Balanced distribution: Similar slice sizes

---

8. Within-Subjects Analysis - jjwithinstats()

Purpose

Compare repeated measures or matched observations with appropriate paired tests.

When to Use

• Pre-post comparisons
• Repeated measures designs
• Matched pairs analysis
• Longitudinal data

Example with Paired Data

# Create example paired data
paired_data <- data.frame(
  Subject = rep(1:20, 2),
  Time = rep(c("Pre", "Post"), each = 20),
  Score = c(rnorm(20, 50, 10), rnorm(20, 55, 10)),
  Group = rep(c("Treatment", "Control"), 20)
)

# Reshape for within-subjects analysis
paired_wide <- paired_data %>%
  tidyr::pivot_wider(names_from = Time, values_from = Score)

within_basic <- jjwithinstats(
  data = paired_wide,
  dep1 = "Pre",
  dep2 = "Post",
  dep3 = NULL,  # No third dependent variable
  dep4 = NULL,  # No fourth dependent variable
  )

within_basic$plot

Statistical Output

• Paired t-test or Wilcoxon signed-rank test
• Effect size (Cohen’s d for paired data)
• Confidence intervals for differences

Interpretation Tips

• Significant change: p < 0.05
• Practical significance: Consider effect size magnitude
• Direction of change: Positive vs negative differences

---

9. Waffle Charts - jwaffle()

Purpose

Display proportional data using a grid of squares for intuitive visualization.

When to Use

• Show parts of a whole
• Display survey results
• Present demographic breakdowns
• Alternative to pie charts

Basic Usage

# Waffle chart
waffle_basic <- jwaffle(
  data = mtcars_clean,
  counts = "hp",
  groups = "cyl",
  facet = "am"
)

waffle_basic$plot

Interpretation Tips

• Visual proportion: Each square represents equal units
• Easy comparison: Squares easier to compare than pie slices
• Pattern recognition: Spatial arrangement shows patterns

---

Choosing the Right Analysis

Decision Tree

# Pseudocode decision tree
if (data_type == "one_continuous") {
  use_jjhistostats()
} else if (data_type == "two_continuous") {
  use_jjscatterstats() # or jjcorrmat() for multiple variables
} else if (data_type == "continuous_by_groups") {
  if (groups_independent) {
    use_jjbetweenstats()
  } else {
    use_jjwithinstats()
  }
} else if (data_type == "categorical") {
  if (display_preference == "bars") {
    use_jjbarstats()
  } else if (display_preference == "pie") {
    use_jjpiestats()
  } else if (display_preference == "waffle") {
    use_waffle()
  }
} else if (data_type == "group_means") {
  use_jjdotplotstats()
}

Data Type Requirements

Analysis	Dependent Variable	Grouping Variable	Sample Size
Histogram	Continuous	Optional categorical	n ≥ 10
Scatter	2 × Continuous	Optional categorical	n ≥ 10
Box-Violin	Continuous	Categorical (2+ levels)	n ≥ 5 per group
Correlation Matrix	2+ × Continuous	Optional categorical	n ≥ 10
Dot Chart	Continuous	Categorical (2+ levels)	n ≥ 3 per group
Bar Chart	Categorical	Optional categorical	n ≥ 5
Pie Chart	Categorical	Optional categorical	n ≥ 5
Within-Subjects	2+ × Continuous	Optional categorical	n ≥ 5 pairs
Waffle Chart	Categorical	Optional categorical	n ≥ 10

Best Practices Summary

Statistical Considerations

1. Check assumptions before interpreting results
2. Report effect sizes alongside p-values
3. Consider multiple comparisons when doing many tests
4. Validate findings with appropriate diagnostics

Visualization Guidelines

1. Choose appropriate plot for your data type
2. Use consistent color schemes across related plots
3. Include informative titles and labels
4. Consider your audience when selecting plot complexity

Workflow Recommendations

1. Start with exploratory plots (histograms, scatter plots)
2. Progress to confirmatory analyses (between-groups, correlations)
3. Use grouped analyses to explore interactions
4. Combine multiple approaches for comprehensive understanding

This gallery provides a complete reference for all jjstatsplot analyses. Each function offers extensive customization options - explore the jamovi interface or function documentation for additional parameters and settings.