Same Test, Different Patterns: How ADHD and Autism Show Up Differently Across IQ Subtests

Discover how ADHD and autism appear differently on cognitive tests and why your FSIQ might be misleading. Read the guide and take the RIOT IQ test!

Dr. Russell T. WarneChief Scientist

Same Test, Different Patterns: How ADHD and Autism Show Up Differently Across IQ Subtests

One of the most practically important things a cognitive assessment can do — and one of the things a single composite score consistently fails to do — is reveal the shape of an individual's cognitive profile rather than just its average level. This matters enormously in the context of neurodivergent conditions, where the characteristic finding is not a uniformly elevated or uniformly depressed profile but a distinctively uneven one.

ADHD and autism are the two most commonly assessed neurodevelopmental conditions in both children and adults, and both are associated with characteristic cognitive profiles on standard IQ batteries. Those profiles are meaningfully different from each other — in which indices are affected, in the magnitude of the effects, and in what the pattern implies about underlying cognitive architecture. Understanding both requires moving past the FSIQ and into the index scores that actually carry diagnostic and interpretive information.

Why the FSIQ Misrepresents Both Conditions

Before breaking down the profiles, it's worth being precise about why the Full Scale IQ is a particularly poor summary statistic for neurodivergent individuals. The FSIQ aggregates performance across all index scores into a single composite, which works reasonably well when a person's indices are relatively consistent with each other. When they're not — when two indices are substantially higher than two others — the FSIQ represents none of those four indices accurately. It becomes, in clinical parlance, an uninterpretable number.

Both ADHD and autism consistently produce substantial inter-index scatter. A neurodivergent individual may have verbal comprehension and fluid reasoning indices in the high-average range alongside working memory and processing speed indices in the low-average or below-average range. The FSIQ that results from combining those four numbers sits somewhere in the middle — a value that accurately describes performance on no single cognitive domain, and that can mislead clinicians, educators, and the individuals themselves about what the person is actually capable of.

Adults with ADHD show significant decrements in subtests with working memory and processing speed demands with moderate to large effect sizes and a higher General Ability Index in comparison with the FSIQ — meaning the GAI, which excludes working memory and processing speed, is consistently a more accurate representation of their reasoning ability than the composite that includes those suppressed indices. The same principle applies to autism, where processing speed depression can pull the FSIQ meaningfully below where fluid and verbal reasoning indices sit. In both cases, the appropriate clinical response is to interpret the index profile, not the composite.

The ADHD Cognitive Profile

The cognitive profile associated with ADHD is one of the most consistently replicated findings in clinical neuropsychology. Across decades of research and multiple versions of the Wechsler scales, the same two indices move consistently: working memory and processing speed are depressed, while verbal comprehension and fluid/perceptual reasoning are preserved.

The most comprehensive recent evidence comes from a 2025 study of 719 children and adolescents — 363 with ADHD and 356 controls — assessed using the WISC-IV. Compared to controls, the ADHD group exhibited significantly lower scores in the Working Memory Index, Processing Speed Index, and Cognitive Proficiency Index, with no significant group differences in Verbal Comprehension or Perceptual Reasoning after controlling for age and sex. This finding is consistent with every major WISC study before it — Fenollar-Cortés et al. 2015, Navarro-Soria et al. 2020, and Thaler et al. 2013 all found the same pattern, though with smaller samples.

The adult data converges on the same conclusion. Adults with ADHD show significant decrements in subtests with working memory and processing speed demands with moderate to large effect sizes, with verbal comprehension and perceptual reasoning holding. In early adulthood, the Processing Speed Index score was notably lower than the other indices of the WAIS-IV in ADHD patients, with PSI serving as a useful predictor of ADHD status — a finding relevant to differential diagnosis.

One important developmental caveat: the working-memory and processing-speed dips are most pronounced in childhood and tend to attenuate in adulthood, where the adult processing-speed reduction may disappear altogether in some analyses. Compensation, practice, and changing test demands all plausibly contribute to this attenuation. The profile is sharper and more diagnostically distinctive in children than in adults — which is one reason why adults with ADHD often find their standardized test profile less obviously "ADHD-shaped" than they expected.

The mechanistic basis for this profile is well-established. Working memory deficits in ADHD are linked to hypoactivation of the prefrontal-parietal network during active maintenance tasks — the same network that holds and manipulates information during digit span and letter-number sequencing subtests. Processing speed deficits are linked to slower and more variable neural response timing, which shows up most clearly on symbol coding and scanning subtests where the bottleneck is rapid, consistent motor output rather than reasoning capacity.

The Autism Cognitive Profile

The autism profile is both older in the literature and more distinctive in its subtest-level pattern. The characteristically uneven profile associated with autism has been described consistently since the earliest systematic cognitive studies of the condition, and the core elements have remained stable across four revisions of the Wechsler scales.

The headline finding is a dissociation between two specific subtests: high Block Design and low Coding scores have been consistently found regardless of variation in intellectual functioning or diagnosis across four versions of the WISC. Block Design — assembling colored blocks to match a target pattern — is a strength that has been documented across the literature since the 1970s. Coding — rapidly copying symbols paired with simple shapes or numbers — is the most reliably depressed subtest in autism, sitting considerably below other performance measures and reflecting the severe processing speed depression characteristic of the condition.

The 2024 meta-analysis by Wilson, published in the Archives of Clinical Neuropsychology, provides the most rigorous quantitative summary of the autism profile. Analyzing data from over 1,800 neurodivergent individuals across 18 studies, the review found that autistic children and adults performed in the typical range for verbal and nonverbal reasoning, but scored approximately one standard deviation below the mean for processing speed and had slightly reduced scores on working memory. The conclusion is unambiguous: autism appears to be associated with a cognitive profile of relative strengths in verbal and nonverbal reasoning and a weakness in processing speed.

The Coding weakness is worth explaining in more detail because it is specific enough to be diagnostically useful. Coding draws on processing speed, visual perception, visuomotor coordination, cognitive flexibility, and sustained attention. Factors such as motor coordination problems and cognitive flexibility difficulties both contribute to the Coding depression in autism — it's not purely a speed problem, but a compound difficulty involving the rapid shifting and motor execution the task demands. Similarly, Comprehension — a verbal subtest asking test-takers to explain social conventions and practical reasoning — is reliably depressed in autism, reflecting the difficulty many autistic individuals have with the kind of intuitive social knowledge the subtest rewards.

The Block Design strength has been explained through the weak central coherence account: autistic individuals tend toward local, detail-focused processing — an ease with mentally segmenting a whole into its parts — rather than global processing. Block Design rewards exactly this style, because the task is solved more efficiently by treating each block's quadrant independently rather than by first holistically perceiving the target pattern and then trying to decompose it. This is a genuine cognitive strength, not a byproduct of deficits elsewhere, and it's important to name it as such.

A 2026 meta-analysis comparing ASD and intellectual disability across Wechsler scales further confirmed that ASD demonstrates a more consistent intellectual profile than ADHD, with comparative studies between ASD and ADHD showing stable patterns in autism and a less distinct or reliable pattern in ADHD. This consistency across instruments and revisions strengthens confidence in the autism profile as a robust finding rather than a measurement artifact.

Where the Two Profiles Overlap and Where They Diverge

Understanding how the two profiles compare is clinically important, particularly because ADHD and autism co-occur at high rates — estimates suggest that 50–70% of autistic individuals also meet criteria for ADHD — and disentangling the cognitive contribution of each condition in a co-occurring profile requires knowing what each looks like independently.

Processing speed is depressed in both conditions. This is the clearest area of overlap, and it contributes to why the two conditions can be difficult to distinguish from processing speed data alone. Both autistic individuals and those with ADHD show PSI scores substantially below their own VCI and fluid reasoning scores. The mechanisms differ: in ADHD the PSI depression is driven primarily by attentional inconsistency and response variability, while in autism it reflects the compound difficulty with rapid visuomotor processing and cognitive shifting described above — but the surface presentation on a PSI score is similar.

Working memory is clearly depressed in ADHD and mildly reduced in autism. The magnitude of the working memory deficit distinguishes the two: in ADHD, WMI depression is a central, defining feature of the profile with large effect sizes, particularly in children. In autism, WMI is slightly reduced but not the primary distinguishing characteristic — the PSI depression is considerably more pronounced relative to reasoning indices.

Verbal and nonverbal reasoning are preserved in both conditions relative to population norms, though autism can show within-profile variability that ADHD typically does not. The Block Design peak and Comprehension depression in autism create a distinctive within-perceptual-reasoning scatter that ADHD does not replicate. In ADHD, fluid and verbal reasoning indices tend to cluster relatively close to each other; in autism, subtest-level scatter within the perceptual domain is more pronounced.

The practical upshot of this comparison: a profile with large WMI and PSI depression alongside preserved VCI and fluid reasoning, and without pronounced within-domain scatter, points more toward ADHD. A profile with severe PSI depression, a pronounced Block Design peak, a Comprehension weakness, and milder WMI reduction points more toward autism. A profile showing both patterns — severe PSI and WMI depression, Block Design peak, Comprehension weakness — is consistent with co-occurring ADHD and autism and warrants clinical evaluation for both.

What This Means for Score Interpretation

The implications for how assessors and test-takers should interpret neurodivergent cognitive profiles are direct. Neither the ADHD profile nor the autism profile produces a FSIQ that accurately represents a person's reasoning ability. In both cases, the GAI — which excludes working memory and processing speed — provides a better estimate of general reasoning capacity when those indices are substantially depressed by condition-specific factors.

This distinction matters for educational placement, workplace accommodations, and clinical decision-making. A person with ADHD whose FSIQ is reported as 105 but whose VCI and fluid reasoning indices both sit at 115 is being systematically underrepresented by the composite. The 105 reflects working memory and processing speed depression driven by ADHD symptoms, not a general reasoning limitation. Similarly, an autistic individual with a FSIQ of 95 and a PSI of 75 alongside VCI and fluid reasoning indices of 105 is carrying a score that misrepresents both their verbal and reasoning strengths and the severity of their processing speed difficulty.

The diagnostic recommendation that follows from this literature is consistent: for neurodivergent individuals, index scores should be interpreted individually, the GAI should supplement or replace the FSIQ as the primary ability estimate when WMI and PSI are substantially depressed, and significant inter-index scatter should be documented and explained rather than averaged away into a single number.

The Takeaway

ADHD and autism each produce characteristic cognitive profiles on standard Wechsler batteries that are meaningfully distinct from each other and from the broader population, but both share one critical feature: they produce inter-index scatter that makes the FSIQ misleading as a primary interpretive anchor. ADHD is characterized by WMI and PSI depression alongside preserved verbal and fluid reasoning, with the profile most pronounced in children and attenuating somewhat in adults. Autism is characterized by severe PSI depression, a Block Design strength, a Comprehension weakness, and relatively preserved verbal and nonverbal reasoning — a profile that has remained stable across four revisions of the Wechsler scales and is now supported by meta-analytic data from over 1,800 participants.

Reading either profile correctly requires looking at the indices separately, understanding the mechanisms driving each pattern, and resisting the pull of the single composite number that flattens the diagnostic signal the profile is carrying.

If you want to understand your own cognitive profile across the indices that these conditions most directly affect — and see your scores reported at the domain level rather than collapsed into a single composite — the RIOT gives you exactly that level of detail.

References

PubMed Central / MDPI Behavioral Sciences. (2025). Cognitive Profiling of Children and Adolescents with ADHD Using the WISC-IV — 719-participant study. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12467670/
PubMed / Journal of Attention Disorders. (2016). Neuropsychological Profiles on the WAIS-IV of Adults With ADHD. https://pubmed.ncbi.nlm.nih.gov/24448224/
PubMed Central. (2025). Neurocognitive Profiles of Early Adulthood ADHD — PSI as ADHD predictor on K-WAIS-IV. https://pmc.ncbi.nlm.nih.gov/articles/PMC11725659/
Cogn-IQ.org. (2026). ADHD and IQ: The Index-Score Gap. https://www.cogn-iq.org/blog/adhd-and-iq/
Oxford Academic / Archives of Clinical Neuropsychology. (2024). Cognitive Profile in Autism and ADHD: A Meta-Analysis of Performance on the WAIS-IV and WISC-V — Wilson 2024. https://academic.oup.com/acn/article/39/4/498/7286382
Springer / Journal of Autism and Developmental Disorders. (2021). Review of Cognitive Characteristics of ASD Using Performance on Six Subtests on Four Versions of the WISC. https://link.springer.com/article/10.1007/s10803-021-04932-x
Cogn-IQ.org. (2026). Autism and IQ: The Cognitive Profile. https://www.cogn-iq.org/blog/autism-and-iq/
MDPI / Psych. (2026). Cognitive Profile of Autism and Intellectual Disorder in Wechsler's Scales: Meta-Analysis. https://www.mdpi.com/2254-9625/16/1/12