CAT™

Halstead Category Test (HCT )

The HCT is the full, computer software version of the Halstead Category Test. This version typically takes 30 to 40 minutes to administer. Although, it is possible that impaired respondents may take up to 95 minutes to complete this assessment.

Adaptive Category Test (ACat)

The ACat, an interactive, computer version of the Cat, uses built-in archival data to compare a subset of responses on items of a subtest to statistically-derived clusters of similar response patterns collected from previous respondents. The ACat then uses a matched statistical group (i.e., a group with similar response patterns to those of the respondent) to accurately predict the respondent’s final score for that subtest.

Russell Revised Short Version (RCat)

The RCat groups the 95 figures of the HCT into 6 subtests. This version of the Category Test was shortened by discarding the memory subtest (subtest VII in the traditional version), reducing in half the number of items from most subtests, and reorganizing subtests V and VI into counting and proportional principles respectively. When multiplied by 2.2, the number of errors across the short and full versions f the test are comparable.

Intermediate Category Test (ICat)

The ICat is designed for children aged 9 to 15. It contains 168 items organized into 6 different subtests. The first two subtests are identical to those found in the adult HCT. The principle of subtest III, as with the adult version, is differential positioning of the figure, but the figures displayed between test versions are not always the same. Similarly, subtests IV and V of the ICat use the proportional principles of the HCT adult subtests V and VI, but the figures displayed are different as is the order of presentation. Finally, subtest VII is a memory subtest that presents figures from the previous subtests.

The test involves presentation of figures on a computer screen. Respondents select keys 1, 2, 3, or 4 on the keyboard to respond to the test items. According to standard Cat protocol, the administrator must remain unobtrusively present during administration of the test. Instructions appear on the screen, but because the test is used to detect cognitive impairment/brain damage, the administrator will provide verbal direction where necessary. The Cat has four computer-adapted versions: the HCT, the ACat, the RCat, and the ICat.

Reliability

A split-half reliability coefficient of .90 or above has been reported repeatedly (Kilpatrick, 1970; Klonoff, Fibiger, & Hutton, 1970; Lopez, Charter, & Newman, 2000; Moses, 1985; Shaw, 1966). The test-retest reliability coefficients for impaired groups tend to increase as performance worsens (r = .72 for schizophrenics, .82 to .96 for respondents with brain damage [Matarazzo et al., 1976]). Investigations focusing on the relationship between the Cat and similar assessments produced only modest correlations.

Validity

The Cat is highly sensitive in detecting cognitive impairment.

The normative data for the four tests are based on the Halstead Category Test—the test upon which all four computer versions are derived.