Coefficient of determination
A common definition is 1 − residual sum of squares / total sum of squares. A value of 0 means the model matches that mean baseline under the stated evaluation convention; 1 means perfect predictions. On test data R-squared can be negative when the model performs worse than the baseline. Its interpretation depends on the model and dataset.
Worked example
If residual squared error totals 20 and total squared deviation from the observed mean is 100, R-squared = 1 − 20/100 = 0.80.
Common mistake
Claiming R-squared must be between 0 and 1 for every model and test set, or treating a high value as proof of causation.
Where it appears in your exams
| Qualification | Exam | Relationship | Period |
|---|---|---|---|
| ACCA | MA | Direct | September 2026–June 2027 |
| FIA | FMA | Direct | September 2026–June 2027 |