Adaptive behavior assessment: how everyday skills get measured

An adaptive behavior assessment measures how well a person actually manages the ordinary demands of daily life, grouped into conceptual, social and practical skills, and it is usually collected by asking people who know the person well to rate hundreds of specific behaviors. It exists because a diagnosis of intellectual disability cannot be made from an IQ score. Both major diagnostic systems require deficits in adaptive functioning as a separate criterion, and the DSM grades severity on adaptive functioning rather than on IQ.

This page covers what adaptive behavior is, why the IQ number was moved out of the diagnostic criteria, which instruments are in use and what kind of measurement they represent, and the informant-disagreement problem that no publisher has solved.

The American Association on Intellectual and Developmental Disabilities defines adaptive behavior as "the collection of conceptual, social, and practical skills that are learned and performed" by people in everyday life. The DSM describes the same three domains with concrete content.

• Conceptual: Language, reading, writing, mathematics, reasoning, knowledge and memory, plus the practical handling of number, money and time.

• Social: Empathy, social judgment, interpersonal communication, gullibility and naivety, and the capacity to make and keep friendships.

• Practical: Personal care, job responsibilities, money management, recreation, safety, and organising school or work tasks.

AAIDD's 12th-edition definition, published in 2021, states that intellectual disability "is a condition characterized by significant limitations in both intellectual functioning and adaptive behavior that originates before the age of 22." That age-22 boundary is worth noticing, because clinical and legal practice long used 18; an independent review of intellectual disability assessment states the same operational threshold. Readers who want the broader distinctions can see our articles on low IQ and intellectual disability and the difference between intellectual disability and learning disability.

DSM-5 removed IQ scores from the diagnostic criteria for intellectual disability and put them in the text instead. The American Psychiatric Association was explicit about why: doing so "ensures that they are not overemphasized as the defining factor of a person's overall ability, without adequately considering functioning levels," and it noted that "this is especially important in forensic cases." Severity is now graded on adaptive functioning rather than on IQ scores alone.

The number did not disappear. The text describes intellectual functioning "approximately two standard deviations or more below the population, which equals an IQ score of about 70 or below," and because a measured score carries error, the manual expresses that as a range of 65 to 75 rather than a line at 70. AAIDD phrases the same idea as "an IQ test score of around 70 or as high as 75." In the 2022 text revision, DSM-5-TR rewrote the passage to say that while a clinician "should not be bound narrowly to the 65-75 IQ score range," the diagnosis "would not be appropriate for those with substantially higher IQ scores." DSM-5-TR also renamed the condition intellectual developmental disorder, keeping "intellectual disability" in parentheses, to align with the ICD-11 term "disorders of intellectual development."

The Supreme Court made the same point in Hall v. Florida (2014). Freddie Lee Hall had scored 71, and Florida treated 70 as a hard bar stopping any further inquiry. The Court held this unconstitutional, writing that the standard error of measurement "reflects the reality that an individual's intellectual functioning cannot be reduced to a single numerical score," and that Florida's rule "bars an essential part of a sentencing court's inquiry into adaptive functioning."

Empirically the two constructs are related but far from redundant. A meta-analysis of 148 samples and 16,468 participants estimated the population correlation between intelligence and adaptive behavior at .51, and found that "the relation between intelligence and adaptive behavior tended to decrease as IQ increased." A smaller study using AAIDD's own scale found full-scale IQ correlating .64 with the conceptual domain, .39 with practical skills and a non-significant .28 with social skills. An IQ score tells you something about everyday functioning and nowhere near enough.

Three instruments dominate. The form structures and age ranges below come from the publishers' own materials, which is worth stating plainly.

• Vineland-3: The Vineland Adaptive Behavior Scales, Third Edition (Sparrow, Cicchetti and Saulnier, 2016, published by Pearson) comes as a Comprehensive Interview, a Parent/Caregiver Form, a Teacher Form and shorter Domain-Level Forms. Pearson lists the interview and parent forms for ages birth to 90 and the Teacher Form for ages 3 to 21. It reports Communication, Daily Living Skills and Socialization domains, plus Motor Skills and an optional maladaptive behavior index.

• ABAS-3: The Adaptive Behavior Assessment System, Third Edition (Harrison and Oakland, 2015, published by WPS) uses five forms spanning birth to 89, including an Adult Form that can be self-reported. It maps directly onto the three diagnostic domains, reporting Conceptual, Social and Practical composites with 11 skill areas beneath them and a General Adaptive Composite above them. WPS states that the instrument is aligned with AAIDD, DSM-5 and IDEA specifications.

• DABS: The Diagnostic Adaptive Behavior Scale, developed by Tassé, Schalock, Thissen and colleagues and published by AAIDD, was built for one narrow job. Item response theory reduced a pool of 260 items to the 75 that carry the most information near the diagnostic cut point, for ages 4 to 21. Published validation reports convergent correlations of .70 to .84 with the Vineland-II and test-retest coefficients of .78 to .95. AAIDD is both the definitional authority and the publisher here, which readers should factor in.

All three are third-party informant rating scales rather than performance tests. Nobody sits the person down and asks them to cook a meal. A respondent who knows them well reports what they typically do without being asked or reminded. This makes adaptive behavior a measure of typical performance, where an IQ test is a measure of maximal performance under standardised conditions, and the difference explains much of what follows.

If the score is somebody's report, then the score depends on who is reporting. The published agreement figures are not reassuring. A study of 222 to 233 autistic children rated on the Vineland-II by both a parent and an education professional found intraclass correlations of .38 for the overall adaptive behavior composite, .39 for communication, .40 for daily living skills, .29 for socialisation and .33 for motor skills, all in the poor-to-fair range, with parents rating higher overall.

The direction of the gap is not even consistent across studies. One study of 99 people on the autism spectrum rated with the ABAS-3 found parents reporting higher general adaptive composites than teachers (means of 68.2 against 65.4). Another, of 115 young people with autism, intellectual disability or global developmental delay, found ABAS-3 parent and teacher scores only moderately correlated, with teachers rating adaptive skills higher. What is consistent is that two informants who both know a child well can produce standard scores that straddle a diagnostic threshold. Our article on inter-rater reliability covers why agreement figures of this size limit what any single score can support.

There is a floor problem too. A 2026 study of the Vineland-3 Domain-Level Teacher Form in a German sample of pupils with intellectual disabilities found the lowest possible domain score reached by 15.8 percent of the sample, and its authors advised "caution when interpreting standard scores in the very low range." When a scale bottoms out, it stops distinguishing between people who differ.

A competent assessment therefore uses more than one informant where possible, treats disagreement as information rather than noise, reports confidence intervals rather than bare numbers, and reads the adaptive behavior profile against the IQ profile instead of averaging them into a verdict. Where the two point in different directions, that discrepancy is usually a reason to gather more observation rather than to pick the score you prefer.

What it cannot do is settle a diagnosis on its own, any more than an IQ score can. Both systems require limitations in intellectual functioning and in adaptive behavior, originating during the developmental period. An adaptive behavior score also says nothing about why a limitation exists: low scores can follow from restricted opportunity, a sensory impairment, or a mental health condition rather than from limited general ability. Readers curious about the cognitive half of the picture can take a professionally developed IQ test, the Reasoning and Intelligence Online Test, remembering that a cognitive score is one input to a clinical judgment and never the judgment itself.

Most often to establish or rule out intellectual disability, since both DSM and AAIDD criteria require deficits in adaptive functioning. It is also used to set intervention goals, document support needs for school and services, and in capital litigation.

The Vineland-3 and the ABAS-3 are the two most widely used comprehensive scales. AAIDD's Diagnostic Adaptive Behavior Scale is narrower by design, built to be most precise near the diagnostic cut point.

Yes, and that combination rules out a diagnosis of intellectual disability. Intelligence and adaptive behavior correlate at roughly .51 across studies, leaving substantial room for a person to score low on one and unremarkably on the other.

Someone who has known the person well over time and across settings, most often a parent, caregiver or teacher. Some forms allow adult self-report. The person being assessed usually performs no tasks.

Not as a criterion. The IQ range sits in the explanatory text as roughly 65 to 75, and DSM-5-TR clarified that a clinician should not be bound narrowly to that range, while noting the diagnosis would not be appropriate at substantially higher scores.

An adaptive behavior assessment is the part of an intellectual disability evaluation that asks whether a person can actually manage daily life. DSM-5 deliberately moved the IQ number out of the criteria and into the text, and grades severity on adaptive functioning, precisely so that no one is diagnosed on a test score alone. The instruments carrying this weight are informant rating scales, and studies of parent-versus-teacher agreement on the Vineland and the ABAS report correlations in the .3 to .5 range with inconsistent direction of bias. That does not make the assessment optional, since the criteria require it. It does mean a single composite from a single informant is the weakest possible basis for a decision this consequential.

1. American Psychiatric Association. (2013). Intellectual disability (DSM-5 fact sheet). psychiatry.org

2. American Psychiatric Association. (2022). Intellectual developmental disorder (intellectual disability) (DSM-5-TR fact sheet). psychiatry.org

3. American Psychiatric Association. (2022). Diagnostic and statistical manual of mental disorders (5th ed., text rev.). doi.org

4. American Association on Intellectual and Developmental Disabilities. (2021). Defining criteria for intellectual disability. aaidd.org

5. Hall v. Florida, 572 U.S. 701 (2014). law.cornell.edu

6. Fletcher, J. M., & Miciak, J. (2024). Assessment of specific learning disabilities and intellectual disabilities. Assessment, 31(1), 53-74. pmc.ncbi.nlm.nih.gov

7. Alexander, R. M., & Reynolds, M. R. (2020). Intelligence and adaptive behavior: A meta-analysis. School Psychology Review, 49(2), 85-110. eric.ed.gov

8. Tassé, M. J., & Kim, M. (2023). Examining the relationship between adaptive behavior and intelligence. Behavioral Sciences, 13(3), 252. pmc.ncbi.nlm.nih.gov

9. Tassé, M. J., Schalock, R. L., Balboni, G., Spreat, S., & Navas, P. (2016). Validity and reliability of the Diagnostic Adaptive Behaviour Scale. Journal of Intellectual Disability Research, 60(1), 80-88. pubmed.ncbi.nlm.nih.gov

10. American Association on Intellectual and Developmental Disabilities. Diagnostic Adaptive Behavior Scale (DABS). aaidd.org

11. Sparrow, S. S., Cicchetti, D. V., & Saulnier, C. A. (2016). Vineland Adaptive Behavior Scales, Third Edition (Vineland-3). Pearson. pearsonassessments.com

12. Harrison, P. L., & Oakland, T. (2015). Adaptive Behavior Assessment System, Third Edition (ABAS-3). WPS. wpspublish.com

13. Moore, H. L., Le Couteur, A., Charman, T., Green, J., Parr, J. R., & Grahame, V. (2023). What is the concordance between parent- and education professional-reported adaptive functioning in autistic children using the VABS-II? Journal of Autism and Developmental Disorders, 53(8), 3077-3091. pmc.ncbi.nlm.nih.gov

14. Stevens, M. A., Selders, K. J., Jeckel, O., Brownfield, V., & Nowell, K. P. (2023). An evaluation of parent and teacher discrepancies on an adaptive behavior measure for youth with autism spectrum disorder, intellectual disability, and global developmental delay. Journal of Autism and Developmental Disorders, 53(7), 2818-2834. doi.org

15. Kirenko, J., Prokopiak, A., & Wodziński, M. (2024). Predictors of adaptive behaviors in individuals on the autism spectrum as assessed by teachers and parents: An analysis based on ABAS-3. Journal of Clinical Medicine, 13(24), 7607. pmc.ncbi.nlm.nih.gov

16. Renner, G., Selmayr, A., Kölbl, S., Dworschak, W., & Scholz, M. (2026). Psychometric properties of the Vineland Adaptive Behavior Scales-Third Edition Domain-Level Teacher Form in a sample of pupils with intellectual disabilities in Germany. Journal of Intellectual Disability Research. pubmed.ncbi.nlm.nih.gov

Hero image: preparing vegetables in a rustic kitchen, by Shixart1985, licensed CC BY 2.0 (creativecommons.org/licenses/by/2.0). Via Wikimedia Commons.