Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: RE: Summarizing Kappa Output

From   Donald Spady <>
Subject   Re: st: RE: Summarizing Kappa Output
Date   Sat, 26 Jun 2010 09:12:01 -0600

Thanks very much. I didn't do that, but your reply stimulated an alternate, somewhat more laborious, but more transparent to me approach.
The job is done.


Martin Weiss wrote:

Try this to get the results into .txt:

webuse rate2, clear

expand 3
set seed 54968
gen byte mygroups=irecode(runiform(),0,.5,.75,1)

tempname hdle
cap erase info.dta

postfile `hdle' str10(var1 var2) int group /* */ kappa prop_agree z using info.dta

qui levelsof mygroups

foreach gr in `r(levels)'{
	kap rada radb if mygroups==`gr'
post `hdle' ("rada") ("radb") (`gr') /* */ (r(kappa)) (r(prop_o)) (r(z)) }

postclose `hdle'

u info, clear
l, noo

outsheet using myfile.txt, replace
!start myfile.txt


-----Original Message-----
[] On Behalf Of Donald Spady
Sent: Samstag, 26. Juni 2010 06:12
Subject: st: Summarizing Kappa Output

It has been a long time since I visited Statalist, but I need some help now.

I have a project with a 'bazillion' kappas and I want to summarize them
Here is my command:
kap var1 var2 if var3 == 'n'

where 'n' can take many values and there are various combinations of var1 and var 2

What I want to be able to do is get output (using the 'return' function I think) so that it looks something like this:
Var1 var2 var3value (e.g. 1) kappa  %Agreement  Z
Var1 var2 var3value (e.g. 2) kappa  %Agreement  Z
Var1 var2 var3value (e.g. 3) kappa  %Agreement  Z

This should allow me to export it to a spreadsheet or a table.
I imagine it could be done by returning the values to a matrix somehow, but I have no experience in this at all
Can I do this? If so, How?

Many thanks.

*   For searches and help try:

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index