IRF_pol: Plot category response curves for polytomous items
In seonghobae/flirt.x32: Flexible Item Response Theory Modeling

Description Usage Arguments Details Author(s) See Also Examples

A function to plot category response curves as a function of a unidimensional ability for unidimensional graded response models (with a cumulative logit link)

1 2	## S3 method for class 'flirt' IRF_pol(alpha=NULL, beta, guess=NULL , inside=NULL, ylim=NULL)

`alpha`	α_i parameter estimates. If `NULL`, α_i=1
`beta`	β_{ir} parameter estimates for category r to item i
`guess`	c_i (guessing) parameter estimates. If `NULL`, c_i=0
`inside`	If `TRUE`, P_i = c_i + \frac{(1-c_i)}{1+ exp(-α_i (θ_p + β_{ir}))} is estimated
`ylim`	the y limits of the plot.

For a single item with three to six response categories.

Minjeong Jeon <jeon.117@osu.edu>

flirt

library(flirt)

## charity data

result4 <- flirt(data=charity, subset=1:100, control=list(minpercent=0.05, nq=5, link="cumulative"))         

beta <- result4@pars[2:14,1]
       
# item response function        
IRF_pol(beta=beta[1:2])  # item 1 with three categories 
IRF_pol(beta=beta[3:5])  # item 2 with four categories