The following is what I have done:
1)##made an data frame:
X=c(1,1,1,1,
1,1,1,2,
1,1,2,1,
1,1,2,2,
1,2,1,1,
1,2,1,2,
1,2,2,1,
1,2,2,2,
2,1,1,1,
2,1,1,2,
2,1,2,1,
2,1,2,2,
2,2,1,1,
2,2,1,2,
2,2,2,1,
2,2,2,2)
X=matrix(X,4,16)
X=t(X)
freq=c(20,2,6,1,9,2,4,1,38,7,25,6,24,6,23,42)
makeDAT=function(PAT,freq){
R=nrow(PAT)
C=ncol(PAT)
DAT=matrix(PAT[1,1:C],C,freq[1])
DAT=t(DAT)
for (i in 2:R){
DAT1=matrix(X[i,1:C],C,freq[i])
DAT=rbind(DAT,t(DAT1))}
return(DAT)
}
DAT=makeDAT(X,freq)
DAT=as.data.frame(DAT)
2)## omit Rash model:
fit1=rasch(DAT, constraint = cbind(length(DAT) + 1,1))
In this model the discrimination paramater is set to 1 for all items.
I would like to see if there is an difference in discrimination and
difficulty
parameters when iam doing the bootstrap method. And if yes, i would like
some
statistics off the difference. So iam trying to do bootstrap
3## omit bootstrap:
model.boot=function(data,indices){
sub.data=data[indices,]
modelrasch=rasch(DAT, constraint = cbind(length(DAT) + 1,1))
coef(modelrasch)
}
rasch.boot=boot(DAT,model.boot,R=1000)
rasch.boot
4## conclusion
I get this:
ORDINARY NONPARAMETRIC BOOTSTRAP
Call:
boot(data = DAT, statistic = model.boot, R = 1000)
Bootstrap Statistics :
original bias std. error
t1* -1.614125150 0 0
t2* -0.071614733 0 0
t3* -0.003301517 0 0
t4* 0.974310609 0 0
t5* 1.000000000 0 0
t6* 1.000000000 0 0
t7* 1.000000000 0 0
t8* 1.000000000 0 0
The big question is what are the t values? I ask R to omit the boot function
to
the coefficients of the Rash model (i.e. coef(raschmodel)->thus
discrimination
and difficulty values). How to interpret? Am i doing something wrong?
Could someone please help me?
Regards
--
View this message in context:
http://r.789695.n4.nabble.com/plz-help-with-rasch-model-tp3298824p3298824.html
Sent from the R help mailing list archive at Nabble.com.