@ManneGerell The following article (sorry for the self promotion) discusses in quite some detail in which cases autocorrelation causes bias: https://t.co/E9oUJN9g2f. It also concludes that lagged y is often not the best way to deal with it
RT @ruettenauer: When do you need spatial models? And which specification should you use? Have a look at my simulation paper, now out in So…
RT @ruettenauer: When do you need spatial models? And which specification should you use? Have a look at my simulation paper, now out in So…
Now has an issue number in Sociological Methods & Research: A systematic comparison of #spatial regression models. The simple SLX does a great job! 🤩 Meanwhile, it has received a best paper award @RC33methodology and @jmwooldridge liked it (or at leas