Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: comparing multiple means with survey data


From   sjsamuels@gmail.com
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: comparing multiple means with survey data
Date   Tue, 23 Jun 2009 12:52:47 -0400

Your syntax is correct.  You don't need the "linearized" option, as it
is the default for -svy: mean-.

However, hypothesis testing is usually not appropriate for finite
population studies.  See:
http://www.stata.com/statalist/archive/2009-02/msg00806.html   If
hypothesis testing is appropriate for your situation , then you should
exclude the finite population correction (fpc) option from your
-svyset- command.

I'm guessing that you also (or only) want to know how different the
means in the categories of var2 are.   Confidence intervals will
provide the answer, and you can keep the finite population correction
in your -svyset- statement if appropriate.

It is poor practice (and cumbersome) to label categories with strings
like "var2a" "var2b". These are  unnecessary as "a", "b", .. have no
descriptive value.  Just make var2 a numeric variable with values 1 2
3 4 5.  Use -label define- and -label values-  to associate the
numeric values with descriptive text.

Assuming that you do that, the easiest  way to to get confidence
intervals for all pairwise differences after -svy: mean- is to write
out the 10 statements

lincom _b[1] - _b[2]
lincom _b[1] - _b[3]
...
lincom _b[4] - _b[5]

For plotting continuous outcomew with groups I recommend -dotplot-
although it will not take weights.

-Steve

On Mon, Jun 22, 2009 at 4:31 PM, Jean-Gael Collomb <jg@ufl.edu> wrote:
>
> Hello -
> I have been struggling to find a way to compare the means of a different categories of one of my variable. I think I have found a way but I wonder if there would be a more efficient way to do it. In the following example, var 2 has five categories (var2a-var2e). Here's teh commands I type (after survey setting the data):
>
> svy linearized : mean var1, over(var2)
> test [var1]var2a = [var1]var2b = [var1]var2c = [var1]var2d = [var1]var2e, mtest(b)
> test [var1]var2b = [var1]var2c = [var1]var2d = [var1]var2e, mtest(b)
> test [var1]var2c = [var1]var2d = [var1]var2e, mtest(b)
> test [var1]var2d = [var1]var2e, mtest(b)
>
> Is there a better way to do this?
>
> Thanks!
>
>
> Jean-Gael "JG" Collomb
> PhD candidate
> School of Natural Resources and Environment / School of Forest Resources and Conservation
> University of Florida
>
>
>
>
> *
> *   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/



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index