So currently I am assessing agreement between two evaluators. They are assessing disease response and the table would like below:
Evaluater 2 CR CRc CRi PR SD PD
Evaluater 1
CR
CRc
CRi
PR
SD
PD
To calculate a weighted Kappa is assigned an Ordinal value to each response: 1, 2, 3, 4, 5, 6. Which sas that uses the following weights as default: 1.00, 0.8, 0.6, 0.4, 0.2.
However, I would really like for CRc and CRi to have the same weights for distance from CR. Does anyone know how I should reorder the ordinal values if even possible, my issue is that they can't have the same ordinal value --- is the only way to combine CRc & CRi? I don't know if that would be clinically meaningful.
What are you basing this statement on: "Which sas that uses the following weights as default: 1.00, 0.8, 0.6, 0.4, 0.2."
What does your current data actually look like is it something like:
Evaluator Value RecordID
There are a number of procedures that will do Kappa coefficients or analysis but we kind of need to know what the raw data looks like to make recommendations.
I can print the weights Sas is using:
ods html;
proc freq data = test5 ;
table xrorres1*bestresp1/agree printkwts ;
test kappa;
weight count;
run;
ods html close;
Data looks like this:
xrorres1 bestresp1 count
CR CR 5
CR CRi 3
CR CRc 2
CR PR 0
CR SD 0
CRi PD 0
etc etc etc
Don't miss out on SAS Innovate - Register now for the FREE Livestream!
Can't make it to Vegas? No problem! Watch our general sessions LIVE or on-demand starting April 17th. Hear from SAS execs, best-selling author Adam Grant, Hot Ones host Sean Evans, top tech journalist Kara Swisher, AI expert Cassie Kozyrkov, and the mind-blowing dance crew iLuminate! Plus, get access to over 20 breakout sessions.
Learn the difference between classical and Bayesian statistical approaches and see a few PROC examples to perform Bayesian analysis in this video.
Find more tutorials on the SAS Users YouTube channel.