Using Putexcel for estat esize

[Kristin Kellner]

Dear All

I have used the putexcel command to successfully export regression results to an excel sheet. Additionally I'd like to export omega-squared from the

estat esize, omega

command for each independant variable in my regression model.

However, I cannot figure out how to do this. So, I run my regression and the effect size command. This is what my effect size output looks like



Within my putexcel command (following the command for exporting my regression results), I have tried

   estat esize, omega
    matrix stats = r(stats)
    matrix o1 = stats[2,1..1]
    putexcel J`row'=matrix(o1), nformat(#.0000)

With this code, for example, I trying to export the omega-squared for "CO" into column J. However, Stata always exports data from my regression table, not from the effect size output.
I feel like I am missing something here, but I cannot get my head around. I was wondering, if it is even possible to use putexcel for commands that do not specify certain variables but instead are connected to a previous command...?

Thank you in advance for your help!
Best
Kristin

[Andrew Musau]

Run either

return list

or

ereturn list

to see if and where results are stored after running a particular command. In this case, Stata stores the results in matrix r(esize).

webuse lbw, clear
regress bwt smoke i.race
estat esize, omega
mat l r(esize)
mat wanted= r(esize)[1..., "omega2".."df_num"]
mat l wanted

Res.:

. estat esize, omega

Effect sizes for linear models

-------------------------------------------
             Source | Omega-Squared      df
--------------------+----------------------
              Model |    .1088457         3
                    |
              smoke |    .0715877         1
               race |    .0806144         2
-------------------------------------------
Note: Omega-Squared values for individual model terms are partial.

.
. mat wanted= r(esize)[1..., "omega2".."df_num"]

.
. mat l wanted

wanted[3,2]
            omega2     df_num
  Model  .10884568          3
  smoke  .07158773          1
1b.race  .08061444          2

.

[Kristin Kellner]

It worked! Thank you so much, Andrew!