No, the count should be 208. You create indicator variables (taking on
values 0 or 1 only) and then summarize them, so the mean of the 0/1
variable is the proportion of individuals who have the variable equal to 1.
The sample over which the mean is computed includes the 0s and the 1s.
The count will bigger than the relevant subpopulation when you use the svy
commands, since svymean x, subpop(repcell1) uses all the data, not just
the people with repcell1==1, for better variance estimates. Read help svy
and linked resources for more info.
Original Message
From: jebeling [mailto:[email protected]]
Sent: Friday, August 27, 2004 1:17 PM
To: [email protected]
Subject: st: Why do i get a failure all the time with this message?
I"ve tried to send this to you several times. I get an error message
every time I try, but here is what I wanted to send.
I'm confused by something in svymean and repcat.
Here is what I've done:
I've created some repcat categories using a variable I've created from
the joint cell frequencies on a crosstabulation I've been using, and I
want to use the ci's for each of the joint frequencies. So I did the
following:
. tab cells
cells  Freq. Percent Cum.
+
1  47 22.60 22.60
2  51 24.52 47.12
3  49 23.56 70.67
4  61 29.33 100.00
+
Total  208 100.00
. tab cells, gen(repcell)
cells  Freq. Percent Cum.
+
1  47 22.60 22.60
2  51 24.52 47.12
3  49 23.56 70.67
4  61 29.33 100.00
+
Total  208 100.00
. ci repcell1
Variable  Obs Mean Std. Err. [95% Conf. Interval]
+

repcell1  208 .2259615 .0290679 .1686545 .2832686
. ci repcell2
Variable  Obs Mean Std. Err. [95% Conf. Interval]
+

repcell2  208 .2451923 .029901 .1862427 .3041419
. ci repcell3
Variable  Obs Mean Std. Err. [95% Conf. Interval]
+

repcell3  208 .2355769 .029495 .1774279 .293726
. ci repcell4
Variable  Obs Mean Std. Err. [95% Conf. Interval]
+

repcell4  208 .2932692 .0316428 .2308857 .3556527
.
Now shouldn't the samples be 47, 51, 49, and 61 for each of the
confidence intervals I've created ci's for? I used the joint
probabilities found in cells, which is derived from a crosstabulation
for the repcell categories. If anyone can clear me up on this I would
appreciate it.
jon ebeling
*
* 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/
*
* 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/