must specify panelvar; use xtset!

[Theofanis Nikolopoulos]

must specify panelvar; use xtset!

What does it mean? Stata stopped with no reason to regress and show in every effort for regression that result.

[Clyde Schechter]

If you are using any of Stata's -xt- regression commands, or a small number of other commands, you must declare your panel variable using the -xtset- command first. -xtset- will also allow you to declare a time variable, but you do not need to do so unless you are going to also be using lag/lead/difference operators or models with autoregressive structure.

The syntax for -xtset- is very simple. It's just -xtset panelvar- or, optionally, -xtset panelvar timevar-. (Replace the italicized terms with the actual names of the panel variable or time variable.)

[Theofanis Nikolopoulos]

Thank you Prof. Schechter !

[Theofanis Nikolopoulos]

region	reg_dummy	group	b var
MAINLAND	0	yes	0,23
MAINLAND	0	no	0,15
MAINLAND	0	yes	0,2
MAINLAND	0	no	0,05
MAINLAND	0	no	0,06
IONIAN SEA	1	no	0,07
N. AIGEAN	2	no	0,02
C. AIGEAN	3	no	0,02
C. AIGEAN	3	yes	0,19
C. AIGEAN	3	no	0,18
C. AIGEAN	3	no	0,23
C. AIGEAN	3	yes	0,29
S.E. AIGEAN	4	no	0,36
S.E. AIGEAN	4	no	0,09
CRETE	5	no	0,15
CRETE	5	no	0,25
CRETE	5	no	0,11
CRETE	5	yes	0,1

I have that dataset and I need to run some tests. But I cannot run the follow: I need to make an anova test in order to control for var b in relation to reg_dummy and group. The point is that I need to control for 5 times because everytime I need to control for the value 0 in relation to all others. Then, a second run is to control for the value 1 in relation to all others. I tried, but the same results. What is wrong?