I want to check for stationarity among my variables in a N=10/T=20 panel dataset. I am using the -xtunitroot llc- command to conduct the Levin-Lin-Chu test for my dependent variable and my list of about 10 independent variables.
When I run the test on each of the variables, some are stationary and some are not. However, if I add the -trend- option to the ones that aren't stationary, they become stationary. When I add the -trend- option to some of the variables that were already stationary without the -trend- option, they become non-stationary. How do I interpret this? Is it okay to add different options (e.g. trends, lag-lengths, etc.) for different variables? Does this mean that I need to have a trend term in my regression?
Thanks!
Related Posts with Different specifications for unit-root tests for different variables
Spatial shocks: when is collinearity too much?Hi, I am working with a municipality-level dataset. I am analysing the effect of immigration shocks…
Trouble with stata outputs: graphing multiple catvars in a stacked bar format in stata, help with using Floatplot I’m not a data scientist so my jargon is going to be really off here but let me try to break this d…
PPML estimation with bilateral trade agreement dummies.Hello everyone! I am applying PPML estimation to analyze the impact of two trade agreements (both b…
The problem with count if syntax regarding encoded string variablesFrom the dataset hospdd given by STATA (https://www.stata-press.com/data/r17/hospdd.dta), I want to …
Confusion about the example in STATA help file of didregressToday I read the detail description of didregress from STATA. from page 6 and 7, there is a point th…
Subscribe to:
Post Comments (Atom)
0 Response to Different specifications for unit-root tests for different variables
Post a Comment