ABSTRACT
One fundamental benefit from constructing measures of human science variables is that the estimates derived from Rasch procedures are located on an interval scale where the unit value of the scale is maintained at all points along the scale. A visual display of this interval scale, such as Rasch item-person maps, more completely conveys the nature of the relations between the item difficulties and the person abilities. Consequentiy, the construction of mathematically meaningful measurement intervals into what is otherwise merely ordinal data paves the way for a closer investigation of the results. Following the example of the physical sciences, we claim that statistical analyses will be effective only if fundamental measures have been constructed in the first place. Moreover, such measures provide the means by which researchers can focus on the changes that occur over time.
