The Level and Nature of Autistic Intelligence II: What about Asperger Syndrome?
Individuals on the autistic spectrum are currently identified according to overt atypicalities in socio-communicative interactions, focused interests and repetitive behaviors . More fundamentally, individuals on the autistic spectrum are characterized by atypical information processing across domains (social, non-social, language) and modalities (auditory, visual), raising the question of how best to assess and understand these individuals’ intellectual abilities. Early descriptions ,  and quantifications (e.g. ) of their intelligence emphasized the distinctive unevenness of their abilities. While their unusual profile of performance on popular intelligence test batteries remains a durable empirical finding , it is eclipsed by a wide range of speculative deficit-based interpretations. (based on socio-cultural arrogance) Findings of strong performance on specific tests have been regarded as aberrant islets of ability arising from an array of speculated deficits (e.g., “weak central coherence”; ) and as incompatible with genuine human intelligence.
For example, Hobson (, p. 211) concluded that regardless of strong measured abilities in some areas, autistics lack “both the grounding and the mental flexibility for intelligent thought.”
Thus, there is a long-standing assumption that a vast majority of autistic individuals are intellectually impaired. In recent years, this assumption has been challenged by investigations that exploit two divergent approaches —represented by Wechsler scales of intelligence and Raven’s Progressive Matrices— to measuring human intelligence . Wechsler scales estimate IQ through batteries of ten or more different subtests, each of which involves different specific oral instructions and tests different specific skills. The subtests are chosen to produce scores that, for the typical population, are correlated and combine to reflect a general underlying ability. Advantages of this approach include the availability of subtest profiles of specific skill strengths and weaknesses, index scores combining related subtests, and dichotomized Performance versus Verbal IQ scores (PIQ vs. VIQ), as well as a Full-Scale IQ (FSIQ) score. However, the range of specific skills assayed by Wechsler scales is limited (e.g., reading abilities are not included), and atypical individuals who lack specific skills (e.g., typical speech processing or speech production) or experiences (e.g., typical range of interests) may produce scores that do not reflect those individuals’ general intelligence.
In contrast, Raven’s Progressive Matrices (RPM) is a single self-paced test that minimizes spoken instruction and obviates speech production or typicality of experiences . The format is a matrix of geometric designs in which the final missing piece must be selected from among an array of displayed choices. Sixty items are divided into five sets that increase progressively in difficulty and complexity, from simple figural to complex analytic items. RPM is regarded both as the most complex and general single test of intelligence ,  and as the best marker for fluid intelligence, which in turn encompasses reasoning and novel problem-solving abilities , . RPM tests flexible co-ordination of attentional control, working memory, rule inference and integration, high-level abstraction, and goal-hierarchy management , . These abilities, as well as fluid intelligence itself, have been proposed as areas of deficit in autistic persons, particularly when demands increase in complexity , , , .
Against these assumptions, we reported that autistic children and adults, with Wechsler FSIQ ranging from 40 to 125, score an average 30 percentile points higher on RPM than on Wechsler scales, while typical individuals do not display this discrepancy, as shown in . RPM item difficulty, as reflected in per-item error rate, was highly correlated between the autistic and non-autistic children (r = .96). An RPM advantage for autistic individuals has been reported in diverse samples. Bolte et al.  tested autistic, other atypical (non-autism diagnoses), and typical participants who varied widely in their age and the version of Wechsler and RPM they were administered; autistics with Wechsler FSIQ under 85 were unique in having a relative advantage on RPM. Charman et al.  reported significantly higher RPM than Wechsler scores (FSIQ and PIQ) for a large population-based sample of school-aged autistic spectrum children. In Morsanyi and Holyoak , autistic children, who were matched with non-autistic controls on two Wechsler subtests (Block Design and Vocabulary), displayed a numeric, though not significant, advantage within the first set of Raven’s Advanced Progressive Matrices items.
The nature of autistic intelligence was also investigated in an fMRI study . Autistics and non-autistics matched on Wechsler FSIQ were equally accurate in solving the 60 RPM items presented in random order, but autistics performed dramatically faster than their controls. This advantage, which was not found in a simple perceptual control task, ranged from 23% for easier RPM items to 42% for complex analytic RPM items.
Autistics’ RPM task performance was associated with greater recruitment of extrastriate areas and lesser recruitment of lateral prefrontal and medial posterior parietal cortex, illustrating their hallmark enhanced perception .
One replicated manifestation of autistics’ enhanced perception is superior performance on the Wechsler Block Design subtest, suggesting a visuospatial peak of ability . Even when autistics’ scores on all other Wechsler subtests fall below their RPM scores, their Block Design and RPM scores lie at an equivalent level .
Thus, enhanced occipital activity, superior behavioral performance on RPM, and visuospatial peaks co-occur in individuals whose specific diagnosis is autism, suggesting an increased and more autonomous role of perception in autistic reasoning and intelligence .
But what about individuals whose specific diagnosis is Asperger syndrome? In Dawson et al.’s previous investigations of autistics’ RPM performance, Asperger individuals were excluded. Asperger syndrome is a relatively low-prevalence  autistic spectrum diagnosis characterized by intelligence scores within the normal range (non-Asperger autistics may have IQs in any range). Two main distinctions between the specific diagnosis of autism and Asperger syndrome are relevant to the question of intelligence in the autistic spectrum. First, while their verbal and nonverbal communication is not necessarily typical across development, Asperger individuals do not, by diagnostic definition, exhibit characteristic autistic delays and anomalies in spoken language. While both autistic and Asperger individuals produce an uneven profile on Wechsler subtests, Asperger individuals’ main strengths, in contrast with those of autistics (see ), are usually seen in verbal subtests (count me in) (as illustrated in ; see also ). Although RPM is often deemed a “nonverbal” test of intelligence, in practice typical individuals often rely on verbal abilities to perform most RPM items. (NOTE: I have commented on this in another post, regarding the pre-test tutoring available to students, during which the “rules of the game” are explained. Is this “cheating” in that “fluid intelligence” and not learned procedures, are supposedly being measured?)
Second, at a group level, Asperger individuals do not display the autistic visuospatial peak in Wechsler scales; rather, their Block Design subtest performance tends to be unremarkably equivalent to their FSIQ (see and also ). The question of whether Asperger individuals display the autistic advantage on RPM over Wechsler is thus accompanied by the possibility that the Asperger subgroup represents an avenue for further investigating the nature of this discrepancy. (I am quite baffled at times by my “native” Asperger experience, which is overwhelmingly visual-sensory, but that verbal language is a “go to tool” for translating that experience into “acceptable” form. Very practical! Why does this “arrangement” seem to occur in Asperger’s?)
Our goal was to investigate whether the autistic advantage on RPM is also characteristic of Asperger syndrome and, further, whether RPM performance reveals a fundamental property of intelligence across the autistic spectrum. If the mechanism underlying autistics’ advantage on RPM is limited to visuospatial peaks or to language difficulties disproportionately hampering Wechsler performance, then the advantage should not be found in Asperger individuals. Indeed, as predicted by Bolte et al. , Asperger individuals should perform even better on Wechsler scales than on RPM. If instead the underlying mechanism is more general and versatile, then Asperger individuals should demonstrate at least some advantage on RPM. Preliminary findings have suggested this to be the case. In one recent study, Asperger children (age 6–12) obtained significantly higher raw scores on RPM than did typical children matched on age and Wechsler performance .
For all the “poo-bah” and graphs, go to original paper (and related papers): https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3182210/
Asperger individuals differ from autistics in their early speech development, in having Wechsler scores in the normal range, and in being less likely to be characterized by visuospatial peaks. In this study, Asperger individuals presented with some significant advantages, and no disadvantages, on RPM compared to Wechsler FSIQ, PIQ, and VIQ. Asperger adults demonstrated a significant advantage, relative to their controls, in their RPM scores over their Wechsler FSIQ and PIQ scores, while for Asperger children this advantage was found for their PIQ scores. For both Asperger adults and children and strikingly similar to autistics in a previous study , their best Wechsler performances were similar in level to, and therefore plausibly representative of, their general intelligence as measured by RPM.
We have proposed that autistics’ cognitive processes function in an atypically independent way, leading to “parallel, non-strategic integration of patterns across multiple levels and scales”  and to versatility in cognitive processing .
Such “independent thinking” suggests ways in which apparently specific or isolated abilities can co-exist with atypical but flexible, creative, and complex achievements. Across a wide range of tasks, including or perhaps
especially in complex tasks, autistics do not experience to the same extent the typical loss or distortion of information that characterizes non-autistics’ mandatory hierarchies of processing
Therefore, autistics can maintain more veridical representations (e.g. representations closer to the actual information present in the environment) when performing high level, complex tasks. The current results suggest that such a mechanism is also present in Asperger syndrome and therefore represents a commonality across the autistic spectrum. Given the opportunity, different subgroups of autistics may advantageously apply more independent thinking to different available aspects of information: verbal information, by persons whose specific diagnosis is Asperger’s, and perceptual information, by persons whose specific diagnosis is autism.
One could alternatively suggest that the construct measured by RPM is relative and thus would reflect processes other than intelligence in autistic spectrum individuals. However, a very high item difficulty correlation is observed between autistic individuals and typical controls, as well as between Asperger individuals and typical controls. As previously noted , these high correlations indicate that RPM is measuring the same construct in autistics and non-autistics, a finding now extended to Asperger syndrome.
Therefore, dismissing these RPM findings as not reflecting genuine human intelligence in autistic and Asperger individuals would have the same effect for non-autistic individuals.
The discrepancies here revealed between alternative measures of intelligence in a subgroup of individuals underline the ambiguous non-monolithic definition of intelligence. Undoubtedly, autistics’ intelligence is atypical and may not be as easily assessed and revealed with standard instruments. But given the essential and unique role that RPM has long held in defining general and fluid intelligence (e.g., ),
we again suggest that both the level and nature of autistic intelligence have been underestimated.
Thus, while there has been a long tradition of pursuing speculated autistic deficits, it is important to consider the possibility of strength-based mechanisms as underlying autistics’ atypical but genuine intelligence.