Ja. Naglieri, SUBTEST AND COMPOSITE SCORE COMPARISONS FOR THE DIAGNOSTIC ACHIEVEMENT BATTERY 2ND-EDITION, Journal of psychoeducational assessment, 11(2), 1993, pp. 139-143
Interpretation of the variation in an individual student's subtest and
composite scores on the Diagnostic Achievement Battery (second editio
n) requires the comparison of many values. Although values needed for
significance when one is making pairwise comparisons are provided in t
he test manual, this method does not control adequately for experiment
wise error rate. That is, if five comparisons are made, and each is te
sted at the .05 level, the error rate for the combination of five comp
arisons is much higher than .05. Use of the Bonferroni correction when
one is comparing each score to the mean score is one method that cont
rols the overall error rate during interpretation of an individual's t
est scores. With this method, the differences required for significanc
e when one is comparing an individual's subtest score to the mean of a
ll the subtest scores were computed at the .05 and .01 levels in accor
dance with the procedure presented by Davis and modified by Silverstei
n. Additionally, the same values when one is comparing Composite score
s for an individual were computed, An example of the use of the techni
que is provided.