Coefficient of Determination (r²)
What You'll Learn
You'll discover how r² quantifies the proportion of variance in one variable that can be explained by another, and how it relates directly to the correlation coefficient you've already learned.
The Core Idea
When you calculate Pearson's correlation coefficient (r), you get a measure of how strongly two variables are linearly related. The coefficient of determination, written as r², takes this one step further by answering the practical question: "How much of the variation in my outcome can be explained by this relationship?"
r² is simply the square of the correlation coefficient, and it ranges from 0 to 1 (or 0% to 100%).
What "Variance Explained" Means
Think of it this way: imagine you're trying to predict students' test scores. Some students score higher, some lower—that's the total variance in scores. Now suppose you know students' study hours, and r = 0.7. The r² = 0.49 tells you that 49% of the variation in test scores can be explained by study hours. The other 51% comes from other factors (sleep, prior knowledge, test anxiety, etc.).
The Relationship to Correlation
- If r = 0.9 (strong positive correlation), then r² = 0.81 (81% of variance explained)
- If r = 0.5 (moderate correlation), then r² = 0.25 (only 25% explained!)
- If r = -0.6 (moderate negative), then r² = 0.36 (direction doesn't matter for r²)
Notice that even "moderate" correlations may explain surprisingly little variance. An r of 0.5 sounds decent, but it explains just a quarter of the variation.
Key Takeaway: r² tells you the proportion of variance in one variable that's explained by its linear relationship with another—it's always between 0 and 1, calculated by squaring the correlation coefficient, and helps you assess the practical importance of a relationship.