Sum of Squares: Total, Between, and Within
What You'll Learn
You'll learn how ANOVA breaks down total variance into two distinct components—variance between groups and variance within groups—to determine if group means truly differ.
The Core Idea
Imagine you're measuring the heights of students from three different schools. Not all students have the same height—there's total variation in the data. ANOVA asks: "Is this variation mostly because students from different schools tend to be different heights (between-school differences), or just because individuals naturally vary regardless of school (within-school variation)?"
ANOVA decomposes the total sum of squares (SST) into two parts:
SST = SSB + SSW
Where:
- SST (Total Sum of Squares): Measures how much each observation differs from the overall mean across all groups
- SSB (Between-group Sum of Squares): Measures how much each group mean differs from the overall mean (scaled by group size)
- SSW (Within-group Sum of Squares): Measures how much individuals vary within their own groups
Think of it like a budget: your total spending (SST) equals money spent on different categories (SSB) plus random day-to-day variations (SSW).
Why This Matters for the F-Statistic
The F-statistic you learned about compares the between-group variance to the within-group variance. By calculating these sums of squares first, we quantify how much variation is due to group membership versus random chance. A large SSB relative to SSW suggests real group differences exist.
Key Takeaway: ANOVA decomposes total variance (SST) into between-group variance (SSB) and within-group variance (SSW), letting us test whether group differences explain meaningful variation in the data.