Abstract
In the foregoing four
articles, four cut-score setting methods are reviewed. We deliberately
avoid making a recommendation of one procedure over others. Rather,
we believe that those responsible for establishing performance
standards and setting the cut-scores in a given situation should
select what they consider the best procedure to use. During the
course of completing the evaluations using Berk's (1986) two sets
of criteria, two main issues were identified that we believe require
more research. These issues concern (a) the presentation of normative
and impact data, and (b) test dimensionality, and given multidimensionality,
the related need to determine which of two scoring models-compensatory
or conjunctive- should be used.
Copyright © AJER, the Faculty of Education, and the University
of Alberta, 2006.
Last revised: May 16, 2006.
Designed by G.H. Buck