Why use a Breusch–Godfrey test for residual serial correlation with a regression model?
StatisticsLevel: IntermediateRegression diagnostics
Why use a Breusch–Godfrey test for residual serial correlation with a regression model?
Hints
- Retain original regressors.
- Add residual lags.
Answer
It uses an auxiliary regression including original regressors and lagged residuals. This accounts for the fitted regression structure rather than checking one isolated correlation.
Source and licence: statsmodels — model diagnostics · BSD-3-Clause · answer: Quantoros editorial