Hello,
I am using log linear kernal regression, where both my dependent and main independent variables are ordinal.
Here I am facing issues:
1. if run simple command without bootstrap, it reports standard errors and coefficients but doesn't report p-vale that will help understand significance of variable.
Command: npregress kernel x y
2. When I use bootstrap it stops working and shows error.
Command: npregress kernel, vce(bootstrap, vce(bootstrap, reps(5))
Error: Bootstrap replications (5): xxxxx done
x: Error occurred when bootstrap executed npregress.
insufficient observations to compute bootstrap standard errors
no results will be saved
I have used nodrop at the end also but that too is not working.
Please help me how to get sign and significant in kernal regression and how can I run bootstrap in this case?
I am using log linear kernal regression, where both my dependent and main independent variables are ordinal.
Here I am facing issues:
1. if run simple command without bootstrap, it reports standard errors and coefficients but doesn't report p-vale that will help understand significance of variable.
Command: npregress kernel x y
2. When I use bootstrap it stops working and shows error.
Command: npregress kernel, vce(bootstrap, vce(bootstrap, reps(5))
Error: Bootstrap replications (5): xxxxx done
x: Error occurred when bootstrap executed npregress.
insufficient observations to compute bootstrap standard errors
no results will be saved
I have used nodrop at the end also but that too is not working.
Please help me how to get sign and significant in kernal regression and how can I run bootstrap in this case?
