10.1177/0163278702250082 ARTICLE
Evaluation & the Health Professions / March 2003
Violato et al. / EXPERT JUDGMENT AND CLUSTER ANALYSIS
This study compares an expert judgment
process—minimal performance levels (MPL)
using the Nedelsky and Ebel procedures—for
setting cutoff scores for pass/fail on licensure
examinations with an empirical approach—
cluster analysis. Data from all three compo-
nents of the Canadian Standard Assessment in
Optometry (CSAO) examinations (knowl-
edge, clinical judgment, and clinical skills)
from 243 candidates were obtained. Results
indicate that for the written components of the
exams employing the Nedelsky method of
MPL setting, there was a mean agreement of
pass/fail of 81% with the cluster analysis
approach on pass/fail categorization. For the
performance exams using the Ebel method,
the mean agreement of pass/fail with the clus-
ter analysis was 93%. Thus the subjective
approaches to setting cutoff scores (i.e., expert
judgment methods) converge with the objec-
tive method (i.e., cluster analysis) of classify-
ing test takers in the same categories.
Keywords: high-stakes examinations; criterion-
referenced testing; minimum per-
formance levels; cluster analysis
A VALIDITY STUDY OF
EXPERT JUDGMENT
PROCEDURES FOR
SETTING CUTOFF SCORES
ON HIGH-STAKES
CREDENTIALING
EXAMINATIONS USING
CLUSTER ANALYSIS
CLAUDIO VIOLATO
University of Calgary and Edumetrics Ltd.
ANTHONY MARINI
University of Calgary and Martek
CURTIS LEE
Royal College of Physicians
and Surgeons of Canada
59
EVALUATION & THE HEALTH PROFESSIONS, Vol. 26 No. 1, March 2003 59-72
DOI: 10.1177/0163278702250082
© 2003 Sage Publications