R help - Jan 2007 - ability estimate with GRM of IRT

Hi my friends,

I have an issue with ability estimates when running GRM of IRT.  I 
have responses from 242 subjects but got 183 ability estimates.  
Below is what I did to get the estimates.

1) I have a csv file "P1.csv" and I imported it into R and loaded 
the "ltm" package by doing:
p1<-read.table("P1.csv",header=TRUE,sep=",")
library(ltm)

2) I created a subset that included columns 2 to 9 for the 
analysis by doing:
s1<-p1[,2:9]

3) I converted the subset into data.frame format by doing:
s1df=data.frame(s1)

4) I checked the descriptive stats for the s1df by doing:
ds1df=descript(s1df)
ds1df

And it was confirmed that 242 subjects' responses were imported.

5)I ran GRM on the s1df dataset by doing:
grm<-grm(s1df, Hessian=T)
grm

6)Finally, I ran the ability estimated by doing:
aes1=factor.scores(grm)
aes1

And I got 183 factor-scores for observed response patterns.  
Besides, the 183 estimates are not for the first 183 subjects of 
the 242 because the response patterns of No 1 on the factor-scores 
list do not match those of No 1 in the s1 dataset and the same 
thing holds true for the rest of the response patterns.

Could you help me find out what the problem is?

Many thanks in advance!

Feng

you need to use the 'resp.patterns' argument of factor.scores(); look 
at ?factor.scores for more info, e.g.,

library(ltm)

fit <- grm(Environment)
factor.scores(fit)
factor.scores(fit, resp.patterns = Environment)


I hope it helps.

Best,
Dimitris

ps, it'd be better not to call your fitted GRM model 'grm'; check

library(fortunes)
fortune("dog")

----
Dimitris Rizopoulos
Ph.D. Student
Biostatistical Centre
School of Public Health
Catholic University of Leuven

Address: Kapucijnenvoer 35, Leuven, Belgium
Tel: +32/(0)16/336899
Fax: +32/(0)16/337015
Web: http://med.kuleuven.be/biostat/
     http://www.student.kuleuven.be/~m0390867/dimitris.htm


----- Original Message ----- 
From: "SHEN,FENG" <fshen at ufl.edu>
To: <r-help at stat.math.ethz.ch>
Sent: Friday, January 19, 2007 5:30 AM
Subject: [R] ability estimate with GRM of IRT

> Hi my friends,
>
> I have an issue with ability estimates when running GRM of IRT.  I
> have responses from 242 subjects but got 183 ability estimates.
> Below is what I did to get the estimates.
>
> 1) I have a csv file "P1.csv" and I imported it into R and loaded
> the "ltm" package by doing:
> p1<-read.table("P1.csv",header=TRUE,sep=",")
> library(ltm)
>
> 2) I created a subset that included columns 2 to 9 for the
> analysis by doing:
> s1<-p1[,2:9]
>
> 3) I converted the subset into data.frame format by doing:
> s1df=data.frame(s1)
>
> 4) I checked the descriptive stats for the s1df by doing:
> ds1df=descript(s1df)
> ds1df
>
> And it was confirmed that 242 subjects' responses were imported.
>
> 5)I ran GRM on the s1df dataset by doing:
> grm<-grm(s1df, Hessian=T)
> grm
>
> 6)Finally, I ran the ability estimated by doing:
> aes1=factor.scores(grm)
> aes1
>
> And I got 183 factor-scores for observed response patterns.
> Besides, the 183 estimates are not for the first 183 subjects of
> the 242 because the response patterns of No 1 on the factor-scores
> list do not match those of No 1 in the s1 dataset and the same
> thing holds true for the rest of the response patterns.
>
> Could you help me find out what the problem is?
>
> Many thanks in advance!
>
> Feng
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 

Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm