Comparing Distributions
You'll learn how to use side-by-side box plots and comparative statistics to systematically analyze differences between two or more groups.
The Core Idea
When you have data from different groups—say, test scores from two classes, or heights from different age groups—you need structured ways to compare them. Simply looking at raw numbers isn't enough. Instead, you combine visual tools (side-by-side box plots) with numerical summaries (comparative statistics) to paint a complete picture of how the distributions differ.
Think of it like comparing two cities: you wouldn't just look at population size. You'd also examine median income, crime rates, temperature ranges, etc. Similarly, comparing distributions means examining multiple features simultaneously.
Side-by-Side Box Plots
Place box plots for each group on the same scale, aligned vertically or horizontally. This visual immediately reveals:
- Center differences: Which group has a higher median?
- Spread differences: Which group shows more variability (wider IQR)?
- Outliers: Does one group have more extreme values?
- Symmetry: Are the distributions similarly shaped or differently skewed?
Comparative Statistics
Support your visual analysis with numerical comparisons:
- Central tendency: Compare means and medians across groups
- Spread: Compare standard deviations, IQRs, and ranges
- Shape: Compare skewness coefficients
- Position: Compare quartiles and percentiles
For example: "Group A has a median of 78 vs. Group B's 85, but Group A's IQR (15) is narrower than Group B's (22), suggesting less variability despite lower center."
Key Takeaway: Effective comparison requires both visual inspection through side-by-side box plots and numerical analysis through comparative statistics—together they reveal differences in center, spread, shape, and outliers between groups.