Kruskal Wallis Test with Frequency Tables

[Lucy Kay]

Does anyone know the code for doing a kruskal wallis test with a frequency table (i.e., one column is the outcome, one column is the outcome score, and the last column is the frequency of people who chose the outcome score). Thank you!

[Carlo Lazzaro]

Lucy:
welcome to this forum.
I'm not clear with what you're after.
Please read the FAQ on how to post more effectively. Thanks.

[Lucy Kay]

I meant, I am trying to do a Kruskal Wallis test for my frequency table. However, Stata won't let me specify [fweight = frequency]. Does anyone know how I am able to specify fweight for the Kruskal Wallis test on Stata?

[Nick Cox]

I would temporarily expand the dataset using your frequency variable. Here's a silly example.

. sysuse auto, clear
(1978 Automobile Data)

. set seed 2803

. gen freq = runiformint(10, 50)

. preserve

. expand freq
(2,129 observations created)

. kwallis mpg, by(rep78)

Kruskal-Wallis equality-of-populations rank test

  +-------------------------+
  | rep78 | Obs |  Rank Sum |
  |-------+-----+-----------|
  |     1 |  54 |  67265.00 |
  |     2 | 243 | 170790.00 |
  |     3 | 893 | 777200.50 |
  |     4 | 553 | 654902.00 |
  |     5 | 284 | 385220.50 |
  +-------------------------+

chi-squared =   274.954 with 4 d.f.
probability =     0.0001

chi-squared with ties =   276.728 with 4 d.f.
probability =     0.0001

. restore

[Lucy Kay]

Oh yay, that worked. Thank you so much!

[Nick Cox]

Good. Please close out your cross-post at https://www.reddit.com/r/stata/comme...r_the_kruskal/