Category Archives: R – 6366

Comparison between ltm and mirt packages for estimating dichotomous IRT models While at AERA 2014, I mentioned to a friend that I was using the ltm package to estimate estimate the IRT models in class. He shared with me that … Continue reading →

Methods of Equating Two Test Forms The code and output below are what we worked through in class. The data sets we used were provided in Blackboard. The first few records of the responses from Forms 1 and 2 are … Continue reading →

Dichotomous IRT models Normal Ogive Model In this section, I show the basic approaches for estimating a normal ogive model. First, I’m going to simulate a five-item data set for us to analyze. library(eRm) irt.NOM <- sim.rasch(500, 5) head(irt.NOM) ## … Continue reading →

The code below produces classical item analysis and was taken from the IRT website of Dr. Brian Habing at the University of South Carolina (http://www.stat.sc.edu/~habing/courses/778S06.html). The chunk of code below defines the items function. items <- function(testdata, correct = T, … Continue reading →

Classical Test Theory Reliability Functions The data used to demonstrate these functions is read in first. data_demo <- read.table(“C:\\Users\\grant_morgan\\Desktop\\Teaching\\EDP 6337 – Psychometric Theory\\Spring 2014\\Data\\data_demo.DAT”, sep = “,”, header = FALSE) The code below contains 3 functions that can aid in … Continue reading →