ConQuest Community
Help desk => Questions and Answers => Topic started by: Georg II on June 29, 2012, 05:43:40 AM

Hallo Ray,
I’m comparing the fit of two models: a 5dimensional within item model and a 4dimensional between item model.
1.Are these two models nested? Can deviance be used to compare them?
In the 5dimensioanl model I have set all the correlations equal to zero while in the 4dimensional model, dimensions are allowed to correlate. Deviance shows that the 4dimensional model fits better.
However, when in the 4dimensional model the correlations among the dimensions are set to zero, the 5dimensional model fits better.
2.Which model fits better, the 4dimensional or the 5dimensional?
I just thought that in order to make all the conditions equal to make a fair comparison I have to set the correlations in the 4dimensional model to zero, too. Besides, the dimensions in the 4dimensional model are test format dimensions. We don’t expect test formats (for example, multiplechoice and gapfilling) to be correlated.
3.Does this make sense?
Thanks
Georg II

Best to post your control file so I can review it. In general a model is a submodel of another model if the submodel can achieved by constraining parameter values of the supermodel. Typically (but not always) the submodel will be a special case of the super model with some parameters fixed at zero