Defining learning
Annotation
disability: what place
does intelligence
testing have now?
Greg O’Brien MB ChB MD FRCPsych FRCPCH
Professor of Developmental Psychiatry, University of
Northumbria.
Correspondence to author at Northgate Hospital, Morpeth,
Northumberland, NE61 3BP, UK.
Why is it important to detect and define learning disability* in
a child, and how can one do so? The answers to the first ques-
tion are, largely, not in dispute. Not only does the recognition
and definition of learning disability contribute to our under-
standing of the basis of problems in a child’s educational,
developmental, and social progress, it also has significant
long-term consequences, especially where the disability is
more severe.1, 2 The diagnosis or medical aetiology of the
child’s learning disability is a significant element of the story
and carries major phenotypic implications, in terms of dys-
morphology, organogenesis, development, and behaviour.3, 4
It is one thing to diagnose, say, Down syndrome, and quite
another to reach an accurate appraisal of the individual
child’s intellect, developmental attainments, and capacities.
In turn, the definition of learning disability is not always
straightforward and is sometimes quite controversial. This is
partly because, traditionally, at least three different approach-
es are used to define learning disability in children, and each
has its inherent attractions and pitfalls. These approaches are:
definition by IQ alone; definition by IQ combined with social
functioning and age at onset;5 and definition by service con-
tact, the so-called ‘administrative definition’.6
Defining learning disability on the basis of IQ alone
Use of intelligence testing alone, or in isolation, for the pur-
pose of identifying learning disability has been widely
employed. Indeed, intelligence testing, whether by Wechsler
Adult Intelligence Test (WAIS)7 or Wechsler Intelligence Scale
for Children (WISC)8 was, in part, devised for this purpose.9
Such an approach relies on the validity and reliability of the
tests in question, which has been seen both as an inherent
attraction and a problem by different authorities.10 This has
been a subject of controversy since the introduction of intelli-
gence testing at the beginning of this century. Critics of intelli-
gence testing have emphasized its shortcomings (e.g. the
study by Barnett11). Test results in one child can vary accord-
ing to mood, motivation, and fatigue, while the tests them-
selves show prominent rehearsal/learning effects, generally
assume a degree of literacy, and are largely framed to suit
*UK usage. US usage: mental retardation
570 Developmental Medicine & Child Neurology 2001, 43: 570–573
Western cultural requirements, with consequent immense
implications for interpretation. These themes emphasize the
limitations of certain approaches to intelligence testing, partic-
ularly in mixed-race populations, and where applied to individ-
uals who have not received appropriate education. According
to proponents of this argument, to employ intelligence testing
to identify a low-IQ group according to a statistically derived
cut-off point is arbitrary, and largely unhelpful. For these rea-
sons, policies of mass application of generic intelligence tests
within large and diverse populations such as school children
have largely fallen out of favour. But those who have played
down the value of intelligence testing by emphasizing the ‘arbi-
trary’ nature of cut-off scores, seem to have forgotten their
derivation and the intention of the original architects of the
tests. It is crucial to bear in mind that cut-off scores identify
lower functioning groups, and that this is based on both a theo-
retical consideration of distribution of intelligence in the popu-
lation as well as a century of experience.12 Also, while not even
the most strident advocates of intelligence testing would sup-
port a return to its mass application, the increasing availabili-
ty of specialized approaches such as Raven’s Matrices13 and
Kauffman Battery14 offer a pragmatic solution to the chal-
lenges which present to testing in a variety of circumstances
and populations. However, the identification of significant
impairment of intelligence through intelligence testing has
long been held by many to have long-term consequences.15
Recent work, in which low IQ in childhood has been reem-
phasized as one major vulnerability factor for subsequent
delinquency, suggests that IQ alone may serve as a useful
marker to identify those with learning disability and that this
in turn may be important as a predictor of subsequent social
adjustment, in addition, of course, to other social and family
factors.16, 17
If IQ is to be used, the cut-off point employed in most
approaches is that at 2 standard deviations below the mean of
100, that is, at score 70. This is the internationally accepted
standard, adopted within the World Health Organization
(WHO) definition of mental retardation. Two other possibili-
ties are frequently adopted. The first is to take those scoring 1
standard deviation below the mean, that is 85. While rarely
being employed systematically, this cut off has some relevance
for special education, where a score of this magnitude corre-
sponds to ‘mild learning difficulty’.18 The other possibility is
to concentrate on those with IQ below 50, the group who
were previously described as ‘the pathological group’, because
of the high prevalence of identifiable pathological causes of
disabling syndromes and brain disorders in that population.19
Selection of the latter group for study is of interest, for this is a
population with an established prevalence and set of needs.20
It has long been established that any individual who falls into
this group is substantially socially disadvantaged in terms of
capacity for independent functioning, and that problems are
likely to endure.1 While recent genetic research gives some
support to this concept of a separate, pathological group of
individuals with very low intelligence,21, 22 a growing body of
evidence emphasizes that polygenic influences are apparent
at all levels of intelligence.3, 23
The consensus remains that the limitations of an approach,
which defines learning disability by intelligence testing
alone, override its attractions. Few authors support the con-
tention that intelligence testing alone is a sufficient or practi-
cal approach to the definition of a population of people with
learning disability, but fewer still would contend that intelli-
gence testing has no role to play here.
Defining learning disability by IQ, social functioning, and age
at onset – the WHO approach
The diagnostic system employed by WHO in respect of learn-
ing disability has three crucial elements: IQ, social functioning,
and age at onset.24 The IQ element is the conventional cut-off
score of 70, within which mild, moderate, severe, and pro-
found subgroups are further sub-classified into mild learning
disability, corresponding to IQ under 70; moderate, IQ under
50; severe, IQ under 35, and profound, IQ under 20. Onset is
said to be ‘within the developmental period’. Not being more
rigorously defined than this, some authorities, notably the
American Association for Mental Retardation, have definition
systems in which any case with onset up to 18 years can be
included.25 However, the majority of instances of acquired
brain damage in mid childhood or later (where islets of normal
functioning are common, and memory problems figure high-
ly), are distinct from pre- and perinatal cases in terms of both
nature and outcome of disability.26, 27 The customary approach
is, therefore, to include pre- and perinatal onset individuals,
plus those with onset in early childhood.28, 29
The inclusion of social functioning in the WHO definition of
learning disability is an important element of this approach.
The principle is to distinguish between ‘impairment, disabili-
ty, and handicap’,5 and to stress the need for some identifiable
social incapacity to be present, in addition to manifest learn-
ing disability. However, the inclusion of social functioning as
an element of the definition of learning disability has been
the subject of some controversy. Much of this criticism cen-
tres on whether labelling people with learning disability as
having social incapacities might stigmatize the individuals
concerned.30 Another issue is the problem of social incapacity
and its social context. For, any individual with a given level of
learning disability might well cope and function sufficiently in
an undemanding social situation, while that same individual
would struggle within a different setting.31 There have been
further concerns that social functioning will inevitably be
context dependent, rather than subject to some broadly
applicable measurement system.32, 33
The issue of definition and measurement of social function-
ing is resolved in recent guidelines on the application of the
WHO ICD-10 Guide to Classification in Mental Retardation,29
which recommend the use of some standard scale, notably the
Vineland Scale of social adjustment.34 The use of this instru-
ment yields an age-standardized quotient, which serves as a
reliable and valid indicator of capacity for independent social
functioning.35 The combination of standardized intelligence
testing, age at onset within the developmental period, and cali-
brated assessment of social functioning, identifies a popula-
tion which is accepted by most authorities as a reliable and
valid indicator of learning disability.
Defining learning disability by service contact – the
‘administrative definition’
An alternative approach to the definition of learning disabili-
ty, which has some widespread support, is the so-called
‘administrative definition’. According to this tradition, those
individuals in contact with learning disability services consti-
tute the best-fit population of people with learning disabili-
ty.36 In the case of children, this population would comprise
those receiving special education for severe learning difficul-
ty. There is a potential set of logical problems here. First of
all, services vary from place to place. Different services,
including schools, do not employ similar or even equivalent
entry criteria.37 Even where they do, entry into special educa-
tion is not a straightforward matter (nor should it be),
depending on parental approval, availability of services, and
other local factors. Furthermore, individual services change
over time. As service models are developed, refined, and
evaluated according to changing local and national policies
and perceived needs, the entry criteria for any one service
will inevitably change. It is therefore most unlikely that any
population identified by this method will be representative
of the total population of people with learning disability.38
Does an ‘administrative definition’ of learning disability
have any advantages? It is likely that the population defined
will be one, which fulfils certain recognizable criteria, being a
smaller group within the total population of children with
learning disability.39 Being readily identifiable in such a way
offers some bonuses, both in terms of potential ease of fol-
low-up, and documentation prepared, particularly up to
point of entry into special education.40 However, the short-
comings of the ‘administrative definition’ of learning disabil-
ity are striking, and include: (1) the extent to which the sample
may not be representative of the learning disability popula-
tion, being skewed towards more severe disability; (2) the
variability of populations identified by such means, varying
both longitudinally and geographically; (3) the immense diffi-
culties which follow from the above, with regard to replication
of research findings, and generalizability of local data for such
purposes as service planning.
New directions in the definition of learning disability
All of these three traditional approaches have their critics.
Definition by IQ, definition according to WHO disability/
impairment/handicap, and definition according to service
contact are all said to share certain limitations. First, insuffi-
cient attention is said to be given to the individual’s actual
social context.41, 42 Moreover, these approaches are seen in
many quarters as likely to label adversely and stigmatize the
individuals concerned.43 For these reasons an increasingly
Annotation 571
convincing body of opinion is more concerned to stress the
positive, functional ability of individuals, rather than their dis-
abilities.44 In line with this thinking, efforts are being made to
incorporate both positive, ability-based elements into the defi-
nition of learning disability, and also to include consideration
of social context and environment.45 Some preliminary efforts
in this direction were to be found in the WHO 1996 publication
ICD-10,29 where one of the five axes employed is a general
measure of ability. This brief (75 page) document did not attain
the status of an official WHO publication. However, in 1997, a
more radical WHO document, ICDH-2: The International
Classification of Impairments, Activities and Participation46
was launched in pilot form for field trials. Here, the emphasis is
changed entirely, and the individual with a disability is consid-
ered in as wide a context as possible. A host of aspects of
lifestyle and situation are included and coded in detail, the
thrust being to record and classify how different personal
impediments and social influences serve to either enhance or
diminish the activities of which the individual is capable, that in
turn affect participation in society. This approach is intended to
foster a new style of holistic thinking and research about dis-
ability. It is likely to succeed in this aim, although it is thought by
some to be somewhat over-detailed and cumbersome for gen-
eral use. Regarding its use in respect of people with learning
disability, the shift towards a more positive and inclusive style
of thinking is certainly to be welcomed. However, it is impor-
tant to note that this document makes no attempt to provide an
operationalized definition of learning disability. On the con-
trary, it is stated specifically that the ICIDH-2 should be used in
conjunction with the ICD-10 definitions. Interestingly, in this
most forward thinking of classification systems, it is acknowl-
edged that the traditional approach to the definition of learn-
ing disability per se remains a crucial foundation in our
thinking and practice. These new insights challenge our think-
ing and widen our perspectives, but they do not negate our
established ideas on definition.
As yet, it is uncertain what the impact of these shifts in
emphasis might be. Certainly, more emphasis on the poten-
tials of individuals with learning disabilities should facilitate
the realization of that potential through maximal develop-
ment of their abilities. Similarly, acceptance of the influence
of the social environment is crucial to the understanding of
any ability or behaviour, not just the situation of the person
with the disability. But it is uncertain how easily these
insights will improve our definition of learning disability. It is
difficult to envisage how some operationalized measure of
social environment might be incorporated into our defini-
tion system. In any event, there is now increasing acknowl-
edgement of the extent to which environmental influences
operate, and in a sense ‘define’ learning disability.
Summary
In summary, therefore, when we endeavour to make any
statements on the presence and/or degree of learning disabil-
ity in children, we need to consider carefully how we define
our population. Intelligence testing, or some similar psycho-
metric procedures should be available, but with two substan-
tial caveats. First, the results of the tests must be taken in the
context of the respective test situations, and any relevant wider
child/environmental factors. In practice, where testing is car-
ried out by a chartered psychologist according to standard pro-
tocols, these considerations are dealt with, and will be cited
572 Developmental Medicine & Child Neurology 2001, 43: 570–573
in the commentary on the test results. Second, intelligence
testing, while helpful, is not sufficient of itself for definition.
There must be evidence for early onset in the developmental
period. Here, it is surprising in clinical practice how often the
evidence is far from clear; population migration and resultant
variations in practice and documentation are contributory fac-
tors here. However, all indications are that evidence of early
onset may become even more important in the near future, in
the face of these latter secular trends. In addition to intelli-
gence testing, therefore, some measure of social functioning
should be made, such as a Vineland assessment.34 It should be
noted, however, that many of the same provisos that apply to
intelligence testing apply here. While the combination of the
two is quite robust, just as intelligence assessment alone is
insufficient, the same applies to assessment of social function-
ing. Furthermore, the complexities of definition by service
received, for example by school attended, must be considered
carefully. To ignore a child’s reception into special education is
to miss important information, whether in terms of the pre-
ceding background factors or the resultant impact of experi-
ence at the school. But to identify or classify a child according
to type of special education received, and to then go on to
make prognostic statements on that basis, is far from satisfac-
tory. Finally, there is a need to be clear about what is meant by
defining learning disability. Social context and obstacles to
integration may in a sense ‘define’ the individual’s situation.
But an operationalized definition of learning disability, includ-
ing the consideration of intelligence, remains vital.
Accepted for publication 20th December 2000.
References
1. Penrose LS. (1972) The Biology of Mental Defect. London, UK:
Sidgwick and Jackson.
2. O’Brien G. (2001) The adult outcome of child mental retardation.
Developmental Medicine & Child Neurology. (Forthcoming).
3. O’Brien G, Yule W, editors. (1995a) Behavioural Phenotypes.
Clinics in Developmental Medicine No. 138. London: Mac Keith
Press.
4. Gillberg C, O’Brien G, editors. (2000) Developmental Disability
and Behaviour. Clinics in Developmental Medicine No. 149.
London: Mac Keith Press.
5. World Health Organization. (1980) International Classification of
Impairments, Disabilities, and Handicaps. Geneva, Switzerland:
World Health Organization.
6. Richardson SA, Koller H. (1996) Twenty Two Years: causes and
consequences of mental handicaps. Cambridge, MA: Harvard
University Press.
7. Wechsler D.(1997) Wechsler Adult Intelligence Test. 3rd edn.
New York: The Psychological Corporation.
8. Wechsler D. (1991) Wechsler Intelligence Scale for Children. 3rd
edn. New York: The Psychological Corporation.
9 Zigler E, Balla D, Hoddap RM. (1984) On the definition and
classification of mental retardation. American Journal of Mental
Deficiency 89: 215–30.
10. Hoddap RM, Zigler E. (1985) Placement decisions and their effects
on the development of individuals with severe mental retardation.
Mental Retardation 23: 125–30.
11. Barnett WS. (1986) Definition and classification of mental
retardation: a reply to Zigler, Balla and Hoddap disorders: from
infancy to adult life. American Journal of Mental Deficiency
91: 111–16
12. Dacey CM, Nelson WM III, Stoe C, Kel J. (1999) Reliability, criterion
validity, and qualitiative comments of the fourth revision of the
Stanford-Binet Intelligence Scale with a young adult population
with intellectual disability. Journal of Intellectual Disability
Research 43: 179–84.
13. Raven JC. (1982) Revised Manual for Raven’s Progressive 39. Richardson SA. (1984) Institutionalisation and
Matrices and Vocabulary Scale. Windsor UK: NFER Nelson. deinstitutionalisation of children with mental retardation. In:
14. Kaufman AS, Kaufman NL. (1983) K-ABC Kaufman Assessment Stevenson HW, Siegal AE, editors. Child Development Research
Battery for Children. Circle Pines MN: American Guidance and Social Policy. Chicago: University of Chicago Press.
Service. 40. O’Brien G, Yule W. (1995b) Methodological issues in
15. Silverstein AB. (1982) Notes on the constancy of IQ. American behavioural phenotypes research. In: O’Brien G, Yule W, editors.
Journal of Mental Deficiency 87: 227–8. Behavioural Phenotypes. Clinics in Developmental Medicine
16. Farrington D. (1994) The twelfth Jack Tizard Memorial Lecture. No. 138. London: Mac Keith Press. p 35–44.
The development of offending and antisocial behaviour from 41. Marks D. (1997) Models of Disability. Disability and
childhood: key findings from the Cambridge Study in Delinquent Rehabilitation 19: 85–91.
Development. Journal of Child Psychology and Psychiatry 42. Tennant A .(1997) Models of disability: a clinical perspective.
36: 929–64. Disability and Rehabilitation 19: 478–9.
17. Fergusson DM, Horwood LJ. (1995) Early disruptive behaviour, 43. Imrie R. (1997) Rethinking relationships between disability,
IQ, and later achievement and delinquent behaviour. Journal of rehabilitation and society. Disability and Rehabilitation
Abnormal Child Psychology 23: 183–99. 19: 263–71.
18. Goacher B, Evans J, Welton J, Wedell K. (1988) Policy and 44. Oliver M. (1998) Disability in health practice and research.
Provision for Special Education Needs – Implementing the 1981 British Medical Journal 317: 1446–9.
Act. London, UK: Cassell. 45. Hutchinson T. (1995) The classification of disability. Archives of
19. Reid AH. (1976) Psychiatric disturbances in the mentally Disease in Childhood 73: 91–3.
handicapped. Proceedings of the Royal Society of Medicine 46. World Health Organization. (1997) ICDH-2: The International
69: 509–12. Classification of Impairments, Activities and Participation.
20. Corbett J. (1979) Psychiatric morbidity and mental retardation. In: Geneva, Switzerland: World Health Organization.
James FE, Snaith RP, editors. Psychiatric Illness and Mental
Retardation. London: Gaskel Press.
21. Gillberg C. (1992) The Emmanuel Miller Memorial Lecture 1991.
Autism and autistic-like conditions: subclasses among disorders of
empathy. Journal of Child Psychology and Psychiatry 33: 813–42.
22. Thapar A, Gottesman II, Owen MJ, O’Donovan MC, McGufin P.
(1994) The genetics of mental retardation. British Journal of
Psychiatry 164: 7–58.
23. Plomin R. (1990) The role of inheritance in behaviour. Science
248: 183–8.
24. World Health Organization. (1992) The ICD-10 Classification of
Mental and Behavioural Disorders. Geneva, Switzerland: World
Heath Organization.
25. Grossman HJ. (1983) Classification in Mental Retardation.
Washington DC: American Association on Mental Deficiency.
26. O’Brien G. (2000) Mental retardation. In: Gillberg C, O’Brien G,
editors. Developmental Disability and Behaviour. Clinics in
Developmental Medicin, No. 149. London: Mac Keith Press.
p 12–26.
27. O’Brien G, Cheeseburgh B. (2000) Traumatic brain damage. In:
Gillberg C, O’Brien G, editors. Developmental Disability and
Behaviour. Clinics in Developmental Medicine, No. 149. London:
Mac Keith Press. p 64–76.
28. Corbett J. (1977) Studies of Mental Retardation. In: Graham PJ,
editor. Epidemiological Approaches in Child Psychiatry. London:
Academic Press.
29. World Health Organization. (1996) ICD-10 Guide to Classification
in Mental Retardation. Geneva, Switzerland: World Heath
Organization.
30. Mittler P. (1979) People Not Patients. London: Methuen.
31. Landesman-Dwyer S. (1981) Living in the community. American
Journal of Mental Deficiency 86: 223–4.
32. Department of Health. (1993) Services for People with Learning
Disabilities and Challenging Behaviour or Mental Health Needs.
London: Her Majesty’s Stationery Office.
33. Evans DW, Hoddap RM, Zigler E. (1995) Mental and chronological
age as predictors of age-appropriate leisure in children with
mental retardation. Mental Retardation
33: 120–7.
34. Sparrow SS, Balla DA, Ciccheti DV. (1984) Vineland Adaptive
Behavior Scale. Circle Pines MN: American Guidance Service.
35. Kamphaus R. (1987) Conceptual and psychometric issues in the
assessment of adaptive behaviour. Journal of Special Education
21: 27–35.
36. Richardson SA, Koller H, Katz H. (1986) A longitudinal study of
males and females in mental retardation services by ages, IQ and
placement. Journal of Mental Deficiency Research 30: 291–300.
37. Department of Health. (1995) Health of the Nation: a strategy for
people with learning disabilities. London: Her Majesty’s
Stationery Office.
38. Mansell J. (1990) The natural history of community mental
handicap teams. In: Brown S, Wistoe G, editors. The Roles and
Tasks of Community Mental Handicap Teams. Aldershot, UK:
CRSP.

Annotation 573