[Carlo Lazzaro]

Anna:
thanks for improvong your way of posting what you typed and what Stata gave you back: it's almost perfect!.
As far as you results are concerned:
- you have two years omitted for different reasons; however, the remainng years do not seem to add anything informative to your results; hence, I would not plug in among predictors;
- the LR test at the foot of the outcome table reaches statistical significance; that means that your choice of going panel is right.

[Anna Solovieva]

Carlo,
Thank you for your help!
As I understand from your comment, the results that I get from random effects regression show me that time effects do not influence my regression. Am I right with this conclusion? I just want to know exactly because I need to explain this.

Kind regards,

Anna

[Carlo Lazzaro]

Anna:
your interpretation of my reply is correct.
Better said, your data do not show any evidence that -i.year- has a statistical significant effect on your -depvar.. This result can be explained considering that a ststistical significant effect does not exist or that your sample size is too lmited to detect one (please, see if interested http://www.bmj.com/content/311/7003/485).

[Anna Solovieva]

Carlo,

Thank you very much for your help! Now, I can move on with my work, as I can understand which model I should use and which are the limitations of the other! I really appreciate your valuable help! I wish you all the best!

Kind regards,

Anna.

[Carlo Lazzaro]

Anna:
I reciprocate the very same to you, too.

[John Adler]

Carlo Lazzaro

Sorry to drop in on a relatively old thread but I had a question. Here you suggested that

your data do not show any evidence that -i.year- has a statistical significant effect on your -depvar.. This result can be explained considering that a ststistical significant effect does not exist or that your sample size is too limited to detect one (please, see if interested http://www.bmj.com/content/311/7003/485).

Would you be able to explain why this is the case? Observing the p-values on 2002, 2004 and 2010 these appear to have some levels of significant effect on the dependent variable.

I am not disagreeing with your response here, but as I face a similar consideration in my own analysis I would be grateful to understand your thought process,

Very best,

John

[Carlo Lazzaro]

John:
please feel totally free to disagree with my replies!
With a bit of hindsight, I should have recommended Anna to test the joint statistical significance of -i.year- via:

testparm(i.year)

[Gabriela Grace]

Hello Carlo, I need a help to interpret the coefficient of my result for time dummy variable like this:

	Robust
FDI_1	Coef.	Std. Err.	t	P>t	[95% Conf.	Interval]
MS_1	.8362559	.0846373	9.88	0.000	.6691876	1.003324
TO	.0084677	.0009072	9.33	0.000	.0066769	.0102584
INFRA_1	.1229908	.0991032	1.24	0.216	-.0726323	.3186139
HC	-1.824411	.3869833	-4.71	0.000	-2.58829	-1.060531

Year
1990	-.2391093	.279986	-0.85	0.394	-.7917831	.3135645

The command that i used is reg FDI_1 MS_1 TO INFRA_1 HC ib2008.Year


anyway i don't know how to preview it as the other previews through via code delimeters

[Carlo Lazzaro]

Gabriela:
are you dealing with cross-sectional or panel data?

[Maya Lim]

Hi everyone! I have a question related to this discussion. I need help in interpreting time dummy coefficients . Time dummy is suggested to be included in the xtabond2 GMM specification. Do I have to interpret the coefficients as well if they are significant? How do I interpret these significant coefficients whether negative or positive?

[Carlo Lazzaro]

Maya:
try:

testparm(i.year)