Internal consistency
As previously noted, items were yoked as to content and virtually equivalent on all indices
(p values, D, alternate form reliabilities). In addition, in the early field test phase, internal
consistency reliabilities were calculated for the scales separately and the values were
consistently within 0.05 of each other. Consequently, to increase the sample size used to
calculate internal consistency from the standardization data, the coefficient was obtained
from only one form. Specifically, most examinees were instructed to complete either Form A
or B because of limitations on their time. Coefficients for the scales, the global scores, and
the ARIK-A composite range from moderately high to high, i.e., from 0.73 (Fluency Scale) to
0.91 (ARIK-A composite). Most scale reliabilities are in the mid-70s and are shown in
Table 2. The standard error of measurement (SEM) is also shown in this table for each scale
and represents an alternative index of test error. These data suggest that examiners can have
the most confidence in the composite score, relative to the scale scores, although all SEM
values are impressive. These values are similar in magnitude to many major published tests
(e.g., see tests' characterizations in Sattler, 2005).