Dear Austin and other STATA Lister,
With your help, I calculated the scalar lcbs & ucbs for the nine est. props. The scalar lcbs & ucbs are the CIs themselves? They are too small numbers which do not include the point estimates, however. How can I get the appropriate CIs for my point estimates with the scalar lcbs and ucbs? Thanks for any suggestion for a novice STATA user. All best, Hong
**** START
5 . svyprop cat1Rr cat2Rr cat3Rr
Survey proportions estimation
+---------------------------------------------------------+
| cat1Rr cat2Rr cat3Rr Obs Est. Prop. Std. Err. |
|---------------------------------------------------------|
| 0 0 0 65 0.202852 0.026513 |
| 0 0 1 29 0.078938 0.020018 |
| 0 1 0 13 0.028383 0.009759 |
| 0 1 1 2 0.003007 0.002141 |
| 1 0 0 5 0.005753 0.003897 |
|---------------------------------------------------------|
| 1 0 1 1 0.003676 0.003676 |
| 1 1 0 2 0.002165 0.001566 |
| 1 1 1 1 0.002011 0.002014 |
| 2 2 2 246 0.673215 0.031302 |
+---------------------------------------------------------+
6 . di "cat1Rr cat2Rr cat3Rr (Lower Bound, Upper Bound)"
cat1Rr cat2Rr cat3Rr (Lower Bound, Upper Bound)
7 . forvalues i=0/2 {
2. forvalues j=0/2 {
3. forvalues k=0/2 {
4. qui gen i`i'j`j'k`k'=(cat1Rr==`i' & cat2Rr==`j' & cat3Rr==`k')
5. qui su i`i'j`j'k`k'
6. if r(max)>0 & r(N)>0 {
7. qui svylogit i`i'j`j'k`k'
8. scalar lcb = invlogit(_b[_cons]-invttail(e(df_r),.025)*_se[_cons])
9. scalar ucb = invlogit(_b[_cons]+invttail(e(df_r),.025)*_se[_cons])
10. di " `i' `j' `k' ( " scalar(lcb) " , "scalar(ucb) " )
> "
11. }
12. }
13. }
14. }
0 0 0 ( .00203944 , .00360586 )
0 0 1 ( .00062643 , .00177758 )
0 1 0 ( .00019229 , .00074874 )
0 1 1 ( .00001 , .00016165 )
1 0 0 ( .0000204 , .00028998 )
1 0 1 ( 6.923e-06 , .00034891 )
1 1 0 ( 7.041e-06 , .00011894 )
1 1 1 ( 3.786e-06 , .00019084 )
2 2 2 ( .0076789 , .01054787 )
**** END
----- Original Message -----
From: "Nichols, Austin" <[email protected]>
Date: Thursday, September 2, 2004 9:00 pm
Subject: RE: RE: st: calculrating confidence Intervals in svyprop statemen ts
> Well, you've got a typo due to wrapping imposed by statalist.
> Lines 10 and
> 11 should be one line. You should really read up ( -whelp egen-
> and -whelp
> tab- ) to figure out how to make a set of indicator variables, and
> then you
> can use -svyci- to get confidence intervals.
>
> Install svyci by typing
> net from http://www-personal.umich.edu/~nicholsa/stata
> net install svyci
>
> -----Original Message-----
> From: [email protected] [[email protected]]
> Sent: Thursday, September 02, 2004 8:35 PM
> To: [email protected]
> Subject: Re: RE: st: calculrating confidence Intervals in svyprop
> statements
>
>
> Dear Austin and other lister,
>
> I used the code from [email protected], which you let me know in
> youre-mail the other day. I believe I didn't make any
> typographical error, but
> STATA didn't calculate CIs and ended up with a caution message (
> invalidname r(199);.
>
> Below is the outputs:
>
>
> **** START
>
> 2. svyprop cat1Rr cat2Rr cat3Rr
>
> Survey proportions estimation
>
> +---------------------------------------------------------+
> | cat1Rr cat2Rr cat3Rr Obs Est. Prop. Std. Err. |
> |---------------------------------------------------------|
> | 0 0 0 65 0.202852 0.026513 |
> | 0 0 1 29 0.078938 0.020018 |
> | 0 1 0 13 0.028383 0.009759 |
> | 0 1 1 2 0.003007 0.002141 |
> | 1 0 0 5 0.005753 0.003897 |
> |---------------------------------------------------------|
> | 1 0 1 1 0.003676 0.003676 |
> | 1 1 0 2 0.002165 0.001566 |
> | 1 1 1 1 0.002011 0.002014 |
> | 2 2 2 246 0.673215 0.031302 |
> +---------------------------------------------------------+
>
>
>
> 3 . di "cat1Rr cat2Rr cat3Rr (Lower Bound, Upper Bound)"
> cat1Rr cat2Rr cat3Rr (Lower Bound, Upper Bound)
>
> 4 . forvalues i=0/2 {
> 2. forvalues j=0/2 {
> 3. forvalues k=0/2 {
> 4. qui gen i`i'j`j'k`k'=(cat1Rr==`i' & cat2Rr==`j' &
> cat3Rr==`k')
> 5. qui su i`i'j`j'k`k'
> 6. if r(max)>0 & r(N)>0 {
> 7. qui svylogit i`i'j`j'k`k'
> 8. scalar lcb = invlogit(_b[_cons]-invttail(e(df_r),.025)*_se
> [_cons])
> 9. scalar ucb =
> invlogit(_b[_cons]+invttail(e(df_r),.025)*_se[_cons])
> 10. di " `i' `j' `k' ( " scalar(lcb) "
> , "
> 11. scalar(ucb) " )"
> 12. }
> 13. }
> 14. }
> 15. }
> 0 0 0 ( .00203944 ,
> ( invalid name
> r(199);
> *** END
>
>
> Did I make any mistake? Any suggestion? Many thanks for your
> invaluablehelp to a novice STATA user. Hong
>
>
>
>
>
>
>
>
>
> ----- Original Message -----
> From: "Nichols, Austin" <[email protected]>
> Date: Wednesday, September 1, 2004 6:16 pm
> Subject: RE: st: calculrating confidence Intervals in svyprop
> statements
> > Note that you report nine categories, and I don't think your CIs
> > will be
> > plausible, given the number of obs and apparent weighting and
> > survey design.
> > Plus your categories are suspect, since if they were 0/2 you
> would
> > have 27
> > categories instead of nine. But mine is not to reason why. Using
> > code from
> > [email protected],
> >
> > di "cat1Rr cat2Rr cat3Rr (Lower Bound, Upper Bound)"
> > forvalues i=0/2 {
> > forvalues j=0/2 {
> > forvalues k=0/2 {
> > qui gen i`i'j`j'k`k'=(cat1Rr==`i' & cat2Rr==`j' &
> cat3Rr==`k')
> > qui su i`i'j`j'k`k'
> > if r(max)>0 & r(N)>0 {
> > qui svylogit i`i'j`j'k`k'
> > scalar lcb = invlogit(_b[_cons]-
> > invttail(e(df_r),.025)*_se[_cons]) scalar ucb =
> > invlogit(_b[_cons]+invttail(e(df_r),.025)*_se[_cons]) di
> "
> > `i' `j' `k' ( " scalar(lcb) " , "
> > scalar(ucb) " )"
> > }
> > }
> > }
> > }
> >
> > gives CIs that are constructed independently and cannot be used to
> > eyeball-test joint hypotheses about proportions. Caveat emptor.
> >
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
