• Identify any significant within-cluster differences

Because the clusters of the DAS assess not only aspects of total cognitive ability, but also separate cognitive skills, it is important to judge whether the clusters themselves are unitary - are the clusters, composed of two subtests - valid measures of the abilities being assessed? Since broad cognitive skills, such as verbal ability, can be measured in numerous ways, it should not surprise examiners that subtests within factors or clusters do often deviate from each other. The DAS Clusters are composed of 2 subtests on the School-Age level, and 2 or 3 subtests on the Preschool level. On each level, the clusters utilize subtests that measure the same broad construct (e.g., Verbal ability) but do so with tasks that are different in their specific task demands. Each subtest can be delineated by one or more "narrow" abilities. These narrow abilities, as defined by Flanagan, McGrew, and Ortiz (2000) and McGrew and Flanagan (1998), assist in the interpretation of the DAS clusters.

To determine the validity of a cluster, examiners must first determine whether the subtests in that cluster differ statistically from one another. As noted above, both the DAS Introductory and Technical Handbook [Table B.1. and Table B.4. (Pp. 290-292)] and the DAS Summary page of the record form include information regarding statistical discrepancy. The values from these sources indicate that, for the Preschool subtests included in the clusters, a difference of 12 and 14 points between the subtests may be considered significant. For the subtests in the School-Age clusters, a difference of 10 to 12 points is needed.

If the clusters appear to be unitary, interpret those clusters as representing broad measures of the separate abilities (e.g., Verbal, Nonverbal Reasoning, Spatial).

• Identify the frequency of any observed significant differences

As with all other comparisons, the base-rate frequency of any obtained within-cluster difference is evaluated for unusualness. In the case of the DAS Preschool subtest comparisons, to reach a level of unusualness, defined as equal to or less than 10% of the sample, a difference of approximately 18 points is need for all comparisons except Verbal Comprehension vs. Naming Vocabulary, where a difference of 13 points is needed. For the School-Age clusters, differences of between 13 and 15 points are necessary.

• If there are differences that are significant and unusual, interpret narrow abilities rather than the cluster