Dear Martin,
Thank you for the posting. I apologize for being unclear and also because I am asking very basic questions.
I am not sure I understand the purpose of this command after let's say a linear regression.
Let's say we have:
xi: regress continvar i.catvar with 3 levels
the regression table will show levels 2 and 3 and compare them with the reference level 1 of the categorical variable. Why would you run the lincom command. I am not sure I understand its purpose. Would you be able to compare level 2 with level 3 with the lincom command?
Regarding the anova post hoc I meant the following:
If I do a one-way ANOVA like below:
oneway continvar catvar, scheffe
I would get the an F-value and p-value which will be telling me whether there is a sig. difference between any of the category levels, whereas the -scheffe- option will do the "post hoc"- pairwise comparisons between the groups 1&2, 1&3 and 2&3 and tell me which groups are actually different.
Are the objectives of the lincom command similar the scheffe option in the above examples? Do I have it all wrong?
Thank you very much.
Nikolaos
--- On Thu, 6/18/09, Martin Weiss <[email protected]> wrote:
> From: Martin Weiss <[email protected]>
> Subject: st: AW: lincom
> To: [email protected]
> Date: Thursday, June 18, 2009, 4:23 PM
>
> <>
>
> What exactly is a "post hoc" test? Apart from that, the
> help file for
> -lincom- does a good job under "Description". What else do
> you want to know
> about the command?
>
> Note that -lincom- works after any estimation command, so
> the -regress- case
> may not be as special as you think...
>
>
>
> HTH
> Martin
>
> -----Ursprüngliche Nachricht-----
> Von: [email protected]
> [mailto:[email protected]]
> Im Auftrag von Nikolaos Pandis
> Gesendet: Donnerstag, 18. Juni 2009 14:46
> An: [email protected]
> Betreff: st: lincom
>
> Hi,
>
> I would like to ask if you would possibly direct me to a
> site/article ect
> that explains in detail the use and meaning of the -lincom-
> command (after
> -regress-).
>
> How is this command related (if it is related) with the
> anova post hoc
> tests.
>
> Thank you very much.
>
> Nick
>
>
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
>
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
