Sunday, June 12, 2005

#19 Rasch Versus Norm Referenced Tests


At IttyBitty Thoughtswe are discussing Rasch analysis and norm referenced tests. It was triggered by Rob's political opposition to the state tests that raised the issue the quality of the state tests. The quality of tests depends on two distinct aspects—their validity or in what the tests say they are testing, and their technical properties or the test calibration techniques. Is a Rasch based test better than a norm referenced test?

Also, we must distinguish between the state's core program of quantitative tests and the ODE math specialist's abortive attempt to fabricate tests of problem solving.

In a nutshell, the state quantitative tests are conceptually bankrupt and technically brilliant. The state has never conducted any validity studies of their tests but Oregon is a national leader in its technical use of the Rasch model. If you want to learn more about the Rasch model, go to our discussion of the Rasch model and follow John's responses.