R² and Residuals Demonstration

Explore how the coefficient of determination (R²) relates to the spread of residuals around the regression line.

R² = 0.00 (No relationship) R² = 0.50 R² = 0.99 (Perfect fit)

Actual R²

Correlation (r)

Slope (b)

Intercept (a)

Residual SD

Variance Decomposition

% Explained by X % Unexplained (Residual)

Understanding R² and Residuals

R² (coefficient of determination) represents the proportion of variance in Y that is predictable from X. When R² is high, points cluster tightly around the regression line and residuals (vertical distances from points to the line) are small.

Residuals are the vertical distances between observed values and predicted values (the line). As R² decreases, residual variance increases—the points scatter more widely around the line, making predictions less precise.

Key relationship: Residual Variance = Total Variance × (1 - R²)