I have used R a lot in the past, but never for simulation. I have a code in
SAS for the Graded Response Model (GRM), also known as Samejima's model. This
code simulates an ordinal response, provided item characteristics (A=item
discrimination, BB(G) are thresholds between various categorical responses).
It is a macro file. I am thinking that I can write this as a function, and
call it up inside a simulation code. Here is the SAS code:
%MACRO GRGEN;
DO G=1 TO NCAT-1;
Z=EXP(A*(THETA-BB(G))); PS(G)=Z/(1+Z);
END;
PP(1)=1-PS(1); PP(NCAT)=PS(NCAT-1);
DO G=2 TO NCAT-1;
PP(G)=PS(G-1)-PS(G);
END;
X=RANUNI(-1);
SUMP=0; R(J)=1;
DO K=1 TO NCAT-1;
SUMP=SUMP+PP(K);
IF X>SUMP THEN R(J)=K+1;
END;
%MEND GRGEN;
Now, I am totally unfamiliar to simulation in R. So does anyone have a good
reference I could go to convert this? Or have any suggestions for how to
convert it to R?
My biggest problem is all the loops inside this program. In particular, how
to setup the updating of R(J).
It seems if I built a function for this, I need the item parameters A and
BB's (possibly the NCAT).
Any suggestions?
Thanks for any help or info.
Keith Yang
University of Tennessee
---------------------------------
[[alternative HTML version deleted]]
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
