Abstract:
Measurement is at the heart of any good research project. Although Classical test theory is the most widely used measurement theory, many alternative theories have emerged in the last hundred years. Rasch modeling is one such theory and by taking a different approach to measurement holds the possibility of enhancing measurement development in the field of family science. To test this idea, a common scale of relationship quality, the Relationship Assessment Scale (RAS), was administered and checked for reliability using both classical test theory and Rasch modeling. The scale was shown to be reliable by classical test theory standards, but Rasch modeling showed improvement was possible. The RAS was modified to make it fit the Rasch model and administered a second time to a new sample. The modified-RAS was then tested for reliability using both classical test theory and Rasch modeling. The modified scale was still reliable by classical test theory standards and now reliable by Rasch modeling standards as well. Rasch modeling offers a wealth of information on what works and what does not work in our scale.
