Noting that -csgof- (Michael N. Mitchell, UCLA) does not allow -if-, the
list of possibilities under this heading includes
-mgof- (Ben Jann, SJ)
SJ-8-2 st0142 . Multinomial GOF: Large-sample svy & small-sample exact
tests
(help mgof if installed) . . . . . . . . . . . . . . . . . .
B. Jann
Q2/08 SJ 8(2):147--169
computes distributional tests for discrete (categorical,
multinomial) variables and supports large-sample tests for
complex survey designs and exact tests for small samples
-chitest-, -chitesti- from -tab_chi- (NJC, SSC)
-cagof- in Steve's posting looks like a typo for -csgof-. I'd cavil at
the description in -csgof- that it's a test for comparing percentages.
It's a test for comparing frequencies, not quite the same thing.
Nick
n.j.cox@durham.ac.uk
Steven Samuels
As asked by the FAQ, please give the source of all non-official commands
The following will work.
***************
preserve
keep if variable ==1
cagof ???
restore
***************
But -mgof- (from SSC) is a better command in many ways, including its
ability to accept an -if- clause.
On Jul 22, 2010, at 6:53 PM, Frank Gallo wrote:
I want to carry out a chi2 goodness-of-fit test, but I want to add a
qualifier (e.g., if variable == 1). The csgof command does not allow
this option. I greatly appreciate any suggestions on how to accomplish
this task.
*
* 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/
