The Assessment of Basic Learning Abilities test (ABLA) is a useful tool for choosing appropriate training tasks for persons with developmental disabilities. This test assesses the ease or difficulty with which persons are able to learn six hierarchically positioned discrimination tasks. A visual-visual nonidentity matching prototype task was examined to assess its (a). relation to the ABLA hierarchy, (b). predictive validity, and (c). test-retest reliability. Results from 23 participants with developmental disabilities suggest that visual-visual nonidentity matching is a worthwhile addition to the ABLA test and is positioned in the ABLA hierarchy above Level 4 (visual-visual identity matching) and below Level 6 (auditory-visual discrimination). The prototype visual-visual nonidentity matching task also demonstrated high predictive validity and test-retest reliability.

The Assessment of Basic Learning Abilities (ABLA) comprises six levels of hierarchically ordered visual and auditory-visual discriminations. Scores on the ABLA, Vineland Adaptive Behavior Scales, and the Wechsler Intelligence Scale-Revised were obtained for 30 participants selected using a stratified random sample (i.e., by level of mental retardation) from 499 people served by a state center. Two noteworthy findings were (a) that correlations between the Vineland domains and the ABLA were all significant and ranged between .65 and .73, and (b) no participant performing below ABLA Level 6 was testable on the WAIS-R. Thus, Level 6 can be used as an efficient screen to determine the testability of this population on standardized intelligence tests. Implications for practitioners are discussed.

Assessment and intervention in Autism and related disorders has become a major application area for the field of Behavior Analysis. However, basic behavioral research and research in other applied areas have recently provided new insights into the nature of language and other complex human behavior that although directly relevant to autism may not be widely known due to the volume of currently available literatures. This paper provides a review and synthesis of recent research on conditional discrimination processes and relational responding in humans, research on a possible hierarchy of discriminations as described by the literature on the Assessment of Basic Learning Abilities (ABLA), and the implications of this research for language acquisition in person with autism and related disabilities.

BACKGROUND Prior studies have showed that presentation methods could affect the accuracy of a choice assessment. METHODS In the current study, high- and low- preferred work tasks were identified in nine adults with developmental disabilities. Both tasks were then introduced in pairs within a choice assessment using the actual tasks, pictures of the tasks and spoken descriptions of the tasks. Participants were also given the Assessment of Basic Learning Abilities (ABLA) test that evaluated their discrimination skills. RESULTS For five of the participants, their consistent choice of preferred task was predicted by their discrimination skills as assessed by the ABLA. CONCLUSIONS These findings extend that of Conyers et al. whereby the systematic assessment of discrimination skills could predict the effectiveness of different presentation methods in this population.