rxx values in plot-method

[Jerry Bean]

Hey Phil,

I have a quick question that might have an obvious answer buried somewhere but I just can' t find it. I am using some of your plots in an article and would like to report how the reliability function displayed in the rxx plot is is computed. Would you share the simple formula used to generate these conditional values. I thought about reporting them as 1-SE**2 (squared) but wasn't sure that is the case.

I would also like to thank you effusively for your work on mirt development. It is a great resource for those of us trying to master IRT. 

Jerry Bean

[Phil Chalmers]

Hi Jerry,

In the plot functions, the rxx plot is just understood as $F(\theta) / ( F(\theta) + VAR(\theta)) $, where F(θ) is the Fisher information and VAR(θ) is the variance defined/estimated by the model (for most IRT models in mirt, this variance is fixed at 1 for identification). The interpretation is that some levels of θ are measured better, and so the proportion of variation between the individuals (the VAR component) explains less variability than the Fisher information measurement component. HTH.

Phil

--
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...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.