Problem adding the p-value of a Kruskal-wallis test to a table created with commands "table" and "collect" in Stata 17

[Magda Gamba]

Dear all,

This is my first post in the community (and also learning how to use Stata), so I hope I manage to follow all the rules. I am using the commands "table" and "collect" to build a table in Stata 17. According to a webinar on the topic presented by Chuck Huber (and hoping that I understood it correctly), to add the results of tests (like p-values), one runs the command for the test and then runs "return list" like in the following example:

ttest age, by(highbp)
return list

After this, appears a list of scalars that can be added to the table. My problem is that I need to add the p-value of a Kruskal_Wallis test and when I run "return list," I only obtain the following scalars:

kwallis f1dpi, by(f1dpiquar)

Kruskal–Wallis equality-of-populations rank test

+----------------------------+
| f1dpiq~r | Obs | Rank sum |
|----------+-----+-----------|
| 1 | 972 | 472878.00 |
| 2 | 981 | 1.44e+06 |
| 3 | 960 | 2.34e+06 |
| 4 | 966 | 3.28e+06 |
+----------------------------+

chi2(3) = 3635.579
Prob = 0.0001

chi2(3) with ties = 3635.609
Prob = 0.0001

return list

scalars:
r(chi2_adj) = 3635.609319306581
r(df) = 3
r(chi2) = 3635.579049760407

There is no scalar for the p-value. Is there a way to add the statistical significance to the table if it does not appear as a scalar? Thank you so much for your help.

Magda

[Andrew Musau]

You can calculate this and add to r() using the -chi2tail()- function. However, the p-value is ambiguous here. Does it refer to the \(\chi^2\) test or the Adjusted \(\chi^2\) test? Something for you to think about!

cap prog drop mystats
program mystats, rclass
return scalar p_adj= chi2tail(r(df), r(chi2_adj))
return scalar Chi2_adj =  r(chi2_adj)
return scalar  DF =  r(df)
return scalar p= chi2tail(r(df), r(chi2))
return scalar  Chi2 =   r(chi2)
end

sysuse census, clear
kwallis medage, by(region)
mystats
return list

Res.:

. kwallis medage, by(region)

Kruskal-Wallis equality-of-populations rank test

  +--------------------------+
  | region  | Obs | Rank Sum |
  |---------+-----+----------|
  | NE      |   9 |   376.50 |
  | N Cntrl |  12 |   294.00 |
  | South   |  16 |   398.00 |
  | West    |  13 |   206.50 |
  +--------------------------+

chi-squared =    17.041 with 3 d.f.
probability =     0.0007

chi-squared with ties =    17.062 with 3 d.f.
probability =     0.0007

.
. mystats

.
. return list

scalars:
               r(Chi2) =  17.0408088235294
                  r(p) =  .000693214758395
                 r(DF) =  3
           r(Chi2_adj) =  17.06211085869512
              r(p_adj) =  .0006862559273437

.

[Magda Gamba]

Dear Andrew,

Thank you so much for sharing the code with me. It worked, and now I got the p-value as a scalar.

Kind regards,

Magda