Coefficient of Determination Calculator
Calculate R² from residual and total sums of squares for a model evaluated against the mean baseline.
Formula
R² = 1 − SSE/SST.
R² ≈ 1-(20)÷(100)
How to use
- Enter the residual sum of squares SSE.
- Enter the positive total sum of squares SST.
- Read R²; it may be negative for a poor model.
Examples
- SSE = 20 and SST = 100 give R² = 0.8.
Frequently asked questions
What are SSE and SST?
SSE sums squared prediction errors. SST sums squared deviations from the observed mean, using the same observations.
Can R² be negative?
Yes, if predictions perform worse than the mean baseline. Ordinary least squares with an intercept on its training sample gives 0 to 1.
Does a large R² prove causation?
No. It describes fit and is not adjusted for the number of predictors.
Sources
Updated: