I'm trying to understand:
how is check for stationarity(or lack thereoff) linked to unit root testing. More so the logic of it. i understand the null hypothesis used in adf or kpss but I need the logic
Why do we include constant and trend term in some instances - I somewhat understand including the trend because if there is a trend in the series we wish to see whether the series is stationary along the trend. Is that correct? and then why include constant?
- why must be evaluate autocorrelation between independent variable when using regression on time series data. The assumption of OLS is that errors should be independently distributed, so checking autocorrelation of residuals I can understand, but why check for autocorrelation between independent variables