Re: Change the interval of the outputted abilities in mirt

[Phil Chalmers]

On Mon, Dec 11, 2017 at 12:30 PM, <paulo....@miraeducacao.com.br> wrote:

I'm using `mirt`and `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 don't follow what you mean. Are you talking about estimating the factor scores in a bounded parameter space, or simply just capping values after the unbounded estimates are obtained? Perhaps the "max_theta" argument is what you want? 

 

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)

I didn't find it in the documentation, but is quite common in IRT analysis for interpretation reasons.

--
You received this message because you are subscribed to the Google Groups "mirt-package" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mirt-package+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.