Relief! Today was much better

Phew! I'm feeling much better today. It's funny how I grumble when by myself, but the instant I'm around other students and profs, I perk right up.

Today in class my prof was going to do a demonstration on the data I sent him on my strategy assessment questionnaire (hmm, I think I need to change that title). This short survey assesses a person's strategies that they develop while doing the Tower of Hanoi task.

The purpose of the task is to get all the disks from peg A to peg C following a set of rules. Only one disk can be moved at a time and a larger disk cannot be placed on top of a smaller disk (but the reverse is true, a smaller disk can be placed on top of a larger disk).

The prof did an item analysis on the survey and found that the survey violated some assumptions of classical test theory. Yessss!!! (I'm sorry if you don't know what I'm talking about, just smile and nod.) So now that means that my pitiful cronbach's alpha of .52 is is not an appropriate measure of the task's reliability (Yes! Throw that alpha in the trash can!). But, unfortunately the items had correlated error terms (No! Say it ain't so!!) and the survey didn't appear to be unidimensional (Nooooooooooooo!!!!!!).

But, the prof told me after class that by throwing out a couple of the items, he found that the survey fit a Mokkon scale (Beautiful, I think I'm going to cry tears of joy). A Mokken scale is a special case of IRT in which the items are ordered from least to most difficult, but the items do not have a meaningful interval between them (i.e, can't say that one item is twice as difficult as another).

So all is right with the world...