The main purpose of the irtpp package is to estimate item parameters according to a IRT model,
to begin the paremeter estimation for a test, use the function irtpp
library(IRTpp) tst = simulateTest(model="3PL") ## Calibrate this test with a 2PL model. est = irtpp(tst$test,"1PL")
Now est
holds the return of the estimation procedure
names(est) est$z est$LL
in est$z
there is a named list of the IRT parameters calibrated for this test. It is also easy to visualize all the Item Caracteristic Curves (ICC) of the test, with the function test.plot
, this is a visual resume of the parameters.
test.plot(est$z)
Alternatively , irtpp
offers other options such as displaying loglikelihood based statistics , reading from files or changing the initial values of the estimation procedure (for advanced uses).
## Calibrating the same test under a 3PL model and displaying the AIC and BIC statistics est = irtpp(tst$test,"3PL", loglikflag=T)
The item caracteristic curve of all the items of the test can be drawn to compare between IRT models.
test.plot(est$z)
Or for an specific item:
test.plot(est$z,2)
IRTpp package also supports estimating latent traits of individuals according to IRT methodologies. User the function individual.traits
to estimate the individual latent traits for a test.
zz = parameter.matrix(est$z,byrow = F) th = individual.traits(model="3PL", itempars = zz,method = "EAP",dataset = tst$test, probability_matrix = est$prob_mat) ##The latent traits. hist(th[,ncol(th)],breaks=40)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.