On this page
Read what the calculator computes
The tool minimises squared residuals for supplied equal-weight paired returns and reports slope, intercept, residual scale and defined fit measures. Those outputs summarise the entered sample.
Retain additional conditions
| Task | Required evidence |
|---|---|
| Descriptive fit | Matched observations and fitting criterion |
| Coefficient inference | Error assumptions, sampling and uncertainty calculation |
| Prediction interval | A specified X, model assumptions and prediction uncertainty |
| Causal interpretation | A supported causal design, not just correlation |
Keep the five-period example scoped
Its R²≈97.350993% and zero residual identities do not establish normal independent disturbances. Five hypothetical points also do not provide a measured historical market distribution.
Choose the actual follow-up
Use residual and scatter views to identify questions worth reviewing. A diagnostic picture can reveal a pattern but cannot by itself certify every statistical assumption. Retain source definitions and the inferential method if the task goes beyond fitting.