Steps in Forecasting and Other Issues

Regarding the errors Hall notes in the time series data, he is *most likely* incorrect in regards to:

Incorrect. The mean–variance issue is that it’s not stationary over time, so Hall is incorrect regarding this issue.

Incorrect. The residual errors Hall notes are an issue, as correlation of the residual errors is a problem with time-series analysis.

Correct! Both issues that Hall highlights are incorrect. When the residual errors are correlated, it violates the linear regression assumption that the residual errors are uncorrelated. When the time series changes over time, the regression results are invalid. In both cases, Hall incorrectly states each error.

the residual errors.

mean–variance issues.

both the residual errors and mean–variance issues.