<>
Omitting the -wald- option to -svy:tabulate-, the results do coincide:
*************
webuse nhanes2b, clear
svy: tabulate race diabetes/*, wald */
tabout race diabetes using table1.xls, c(col) f(4 0) /*
*/ replace clab(percent Count) svy nlab(samplesize) /*
*/ stats(chi2) per
*************
Easy answer: -tabout- has no -passthru- for the -wald- option, so it is only
natural for the results to be different...
HTH
Martin
-----Ursprüngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von Rajaram
Subramanian Potty
Gesendet: Donnerstag, 5. November 2009 11:22
An: [email protected]
Betreff: st: Different p-values
Dear all,
I am getting different p-values as well as uncorrected chi-square
value when I run tabout and the SVY:tab why. following is the results
. tabout typo icvi ///
> using "D:\My Documents\table1.xls", c(col) f(4 0) replace clab(percent
Count)svy nla
> b(samplesize) stats(chi2) per
Survey results being calculated
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
.......
Table output written to: D:\My Documents\table1.xls
icvi
Type of PSU 1 2 Total samplesize
percent percent percent
Home 52.6991 55.7778 54.2406 209
Brothel 20.8022 22.3111 21.5576 109
Public place 26.4987 21.9111 24.2018 97
Total 100.0000 100.0000 100.0000 415
Pearson: Uncorrected chi2(2) = 1.1931
Design-based F(1.45, 74.01) = 7.1319 Pr = 0.004
.
end of do-file
. svy:tab icvi typo, wald
(running tabulate on estimation sample)
Number of strata = 1 Number of obs =
415
Number of PSUs = 52 Population size =
415.441
Design df =
51
--------------------------------------------------
| Type of PSU
icvi | Home Brothel Public p Total
----------+---------------------------------------
1 | .2631 .1039 .1323 .4993
2 | .2793 .1117 .1097 .5007
|
Total | .5424 .2156 .242 1
--------------------------------------------------
Key: cell proportions
Wald (Pearson):
Unadjusted chi2(2) = 13.6306
Adjusted F(2, 50) = 6.6817 P = 0.0027
Thanks and regards,
RAJA
*
* 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/
