Autocorrelation & Time-Series Residuals

Analyze dependence in regression errors from time-ordered observations. Prerequisites: linear regression, variance of linear combinations and residual diagnostics. Compute AR(1) correlations, Durbin-Watson statistics and Cochrane-Orcutt transformations. Examine exogeneity, contrast-specific uncertainty, dynamic-regression caveats and conditional forecasting without requiring advanced time-series model selection.

Cover: Photo by Yan Krukau on Pexels.