mirtCAT libraries to obtain student’s trait scores with an unidimensional dichotomic 3 parameters logistic model (in witch the items are already callibrated).
How can I change the trait scale interval outputted by
fscores? More explicitly, given the vector
F1 containing the scores of the
N test takers, how to enforce that every entry of
F1 is a member of a interval whose extremes are a < b?
I’m using the following code to get the abilities:
library(mirt) library(mirtCAT) test <- read.csv(test.csv) par <- read.csv(par.csv) a1 <- c(par[,1]) d <- c(par[,2]) g <- c(par[,3]) par <- data.frame(a1,d,g) mod <- generate.mirt_object(parameters=par,itemtype='3PL') scores <- fscores(mod,response.pattern=test)
Witch produces abilities from minus infinity to infinity. The documentation does not mention this, but is quite common in IRT analysis for interpretation reasons.