Intelligence Testing in Autistic Children Regarded As Very “Low-Functioning”: The Good Surprise

Background:  Different intelligence tests can portray autistic children’s abilities in very different ways. For some individuals, there can be a considerable mismatch between scores on different tests. For example, many non-speaking autistic children simply cannot be evaluated with the Wechsler scales and are therefore considered to be intellectually disabled. Could these autistic children, often described as very “low-functioning,” show unexpected reasoning capacities within the normal range when assessed with tests better suited to accurately measuring autistic abilities? A wide variety of autistics perform better than non-autistics on several visual perceptual tasks, and their performance on some of these tasks correlates with their reasoning abilities. In a group of so-called very “low-functioning” autistic children, can perceptual abilities predict reasoning abilities as well?

Objectives:  To evaluate the intellectual potential of these “un-testable” autistic children and to investigate associations between their perceptual and reasoning abilities.

Methods:  All children aged six to 12 years from a specialized school for intellectually disabled autistics were solicited for the study. Nineteen children participated, among whom 11 had no spoken words, five used fewer than five single words, and three used a few simple two-word phrases. The study involved evaluation with WISC-IV, Raven’s Coloured Progressive Matrices board form (RCPM), and two perceptual tasks on which autistics are known to perform well, the Children’s Embedded Figures Test (CEFT) and visual search (VS). A control group of 14 typically developing children was recruited and matched to the autistic group on chronological age.

Results:  None of the autistic children could reliably complete any of the WISC-IV subtests. However, 16 of 19 autistic children could reliably complete other tasks. Autistic children spontaneously understood the CEFT and VS with no or minimal demonstration, and were similar to the control group in pattern of results despite generally slower response times. For RCPM, 15 autistic children could reliably complete the test following brief training on simple puzzles with a visual display similar to RCPM. Interestingly, eight of the 15 autistic children who completed the RCPM scored in the normal intelligence range, one child scoring well above average (90^th percentile). These eight autistic children were generally faster in CEFT and VS than the seven who scored in the intellectual disability range on RCPM.

Conclusions:  These results show that non-speaking “un-testable” autistic children who are regarded as very “low-functioning” can be evaluated with certain tests that require few or no verbal indications. More importantly, some non-speaking autistic children considered to have an intellectual disability were shown to have intellectual potential in the normal range, which bears implications for educational interventions. Furthermore, a possible association between performance in simple-to-administer perceptual tasks and performance in more complex reasoning abilities could be of particular importance when facing difficulties in the evaluation of non-speaking and/or very young autistic children. Indeed, even if no evaluation seems possible with an autistic child, obtaining a good performance on one of these tasks could suggest hidden potential that should be encouraged.