I would really appreciate it if you could advise me on this question. How to test for autocorrelation when OLS regression contains cross sectional and time series data? My dataset consists of over 700k unique loan observations over the period of 4 months. The dataset has some observations from different countries (not panel data). I am studying the relationship between interest rates and a unique independent variable X with borrower and loan characteristics as control variables. However, I also want to control for macroeconomic variables such as unemployment rate(monthly data) in each country and bond yield(monthly data). My loan data is cross sectional; however, the addition of these 2 macroeconomic variables may change the nature of data. Most of stata tutorials show testing for autocorrelation in time series data only. But my data is not purely time series except for 2 macroeconomic control variables. What command should I use for testing autocorrelation in this type of data?
These videos are so useful ! Thank you so much!!
Thank you very much for all!
Professor, do you have a strategy for addressing and managing the lack of normality assumption pertaining to the residuals?
Thank you so much !!!!
Hi Mike! May I ask why you didn't test for autocorrelation? Isn't that one of the assumptions?
I would really appreciate it if you could advise me on this question. How to test for autocorrelation when OLS regression contains cross sectional and time series data? My dataset consists of over 700k unique loan observations over the period of 4 months. The dataset has some observations from different countries (not panel data). I am studying the relationship between interest rates and a unique independent variable X with borrower and loan characteristics as control variables. However, I also want to control for macroeconomic variables such as unemployment rate(monthly data) in each country and bond yield(monthly data).
My loan data is cross sectional; however, the addition of these 2 macroeconomic variables may change the nature of data. Most of stata tutorials show testing for autocorrelation in time series data only. But my data is not purely time series except for 2 macroeconomic control variables. What command should I use for testing autocorrelation in this type of data?