7.6 Test-level Absolute Fit measures

If a model fits data, the model predictions and observations should be close to each other.Test-level absolute fit measures attempt to identify how well the model reproduces the observed data(Han & Johnson, 2019).

• Test-level absolute fit measures are calculated without comparison to other models.

• Rupp et al. (2010) suggested that absolute fit should always be considered when analyzing test data.

To obtain the test-level absolute fit measure, we often use the full information statistics.

References

Han, Z., & Johnson, M. S. (2019). Global-and item-level model fit indices. Handbook of Diagnostic Classification Models: Models and Model Extensions, Applications, Software Packages, 265–285.

Rupp, A. A., Templin, J., & Henson, R. A. (2010). Diagnostic measurement: theory, methods, and applications. Guilford Press.