This simple macro will compute AC1 (no variance, only the statistic) for
dichotomous outcomes:
********************
* RaterA RaterB *
* a= Yes Yes *
* b= No Yes *
* c= Yes No *
* d= No No *
********************.
DEFINE GWETAC1(a=!TOKENS(1)/b=!TOKENS(1)/c=!TOKENS(1)/d=!TOKENS(1)).
MATRIX.
COMPUTE a=!a.
COMPUTE b=!b.
COMPUTE c=!c.
COMPUTE d=!d.
COMPUTE n=a+b+c+d.
PRINT {a,b,a+b;c,d,c+d;a+c,b+d,n}
/FORMAT='F8'
/RLABEL='B+','B-'
/CLABEL='A+','A-'
/TITLE='Input data'.
COMPUTE p=(a+d)/n.
COMPUTE p1=(2*a+b+c)/(2*n).
COMPUTE e=2*p1*(1-p1).
COMPUTE AC1=(p-e)/(1-e).
PRINT AC1
/FORMAT='F8.3'
/TITLE="Gwet's AC1".
END MATRIX.
!ENDDEFINE.
GWETAC1 a=85 b=5 c=5 d=5.
>
> Would anybody know the syntax to run an AC1-statistic in SPSS?
>
>
>
Marta
--
For miscellaneous statistical stuff, visit:
http://gjyp.nl/marta/=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
