Applying Item Response Theory (IRT) models, such as Rasch models, provides
significant advantages for the development and evaluation of Likert-type items and
instruments (Bond & Fox, 2001). Rasch analysis converts ordinal data into ratioscaled
data and produces item parameters and person parameters that are of a ratio
level of measurement (Bond & Fox, 2001). Rasch-based analyses are also able to test
whether item/scale comparability exists for a given sample by testing if all items are
answered in the same fashion or not. This allows empirical testing of Likert-type
scale assumptions. It also allows for comparisons of students and items on quantitatively
equivalent intervals.