I think Ginger's posting encapsulates the
main moral here, reluctant although
people may be to identify it and accept it.
Ginger's code shows how you can use Stata
to customise output, in terms of what you
want, how you want it, and where you
want to put it. Ginger's choices clearly
are not everybody's: for example,
Ginger likes stars, whereas many statistical
people can't abide them; Ginger is thinking
of output to some part of MS Office,
which many Stata people don't use, etc.
Some Stata users seem to ask questions
like these on the assumption that
there may be -- or should be -- some
program (-niceoutput-, say) which delivers
what you really want so long as you
can understand its syntax. There is
such a program, but it's called Stata.
Or, there are various programs of
the form -myniceoutput- in which
individual authors codify what they
like.
Nick
[email protected]
Ginger A Smith
> Hello Eric,
>
> I'd be interested in hearing from other listers of how they make nice
> tables too.
> Here's the way I do it:
> This example makes a table comparing the means of two groups
> (treatment and
> control). The code below creates column1.out with means and
> column2.out
> with standard errors. The rows are the variables var1 var2
> and var3. I
> can open these files using Excel and easily create the tables
> (especially
> if you use good variable labels). However, its not as nice as
> -outreg- for
> regressions. I'd love to add some code to do a t-test of
> means (control vs.
> treatment), add significance stars and append the columns together.
>
> -------
>
> use data, clear
> sort control
> collapse (mean) var1 var2 var3, by(control)
> xpose, clear format(%9.3f) varname
> outsheet using column1
>
> use data, clear
> sort control
> collapse (sd) var1 var2 var2, by(control)
> xpose, clear format(%9.3f) varname
> outsheet using column2
>
> /* note control is a 0/1 variable, var1-var3 are continuous */
>
> -------
>
> --On Saturday, July 24, 2004 10:59 AM -0400 Eric Jensen
> <[email protected]> wrote:
>
> > I'm a big fan of outreg for exporting regression results to
> Word. I've
> > recently been asked to do a job that had previously been
> done in SPSS,
> > which I won't use (example of why: Stata flagged a perfect
> predictor the
> > first time I reran a probit in Stata, that SPSS had somehow
> allowed into
> > the specification!). SPSS does make nice looking tables,
> though. Is
> > there an equivalent ado file to outreg for crosstabs? Or
> (much to my
> > potential chagrin) is there something new(er) within Stata
> documented in
> > those v8 manuals I've been too cheap to buy?
*
* 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/
