Disco

Journal of Child Psychology and Psychiatry 43:3 (2002), pp 307±325
The Diagnostic Interview for Social and

Communication Disorders: background, inter-rater
reliability and clinical use
Lorna Wing,1 Susan R. Leekam,2 Sarah J. Libby,3 Judith Gould,1 and Michael
Larcombe4
1
Centre for Social and Communication Disorders, UK; 2University of Durham, UK; 3University of East Anglia, UK;
4
University of Plymouth, UK
Background: The Diagnostic Interview for Social and Communication Disorders (DISCO) is a
schedule for the diagnosis of autistic spectrum and related disorders and assessment of individual
needs. It enables information to be recorded systematically for a wide range of behaviours and
developmental skills and is suitable for use with all ages and levels of ability. In addition to helping
the clinician to obtain a pro®le of each individual's pattern of development and behaviour, the DISCO
also enables identi®cation of speci®c features found in autistic spectrum disorders that are relevant
for use with established diagnostic systems. Method: This paper describes the historical background
of the DISCO, outlines its structure and reports the results of an inter-rater reliability study with
parents of 82 children aged 3 to 11 years with autistic spectrum disorder, learning disability,
language disorder or typical development. Results: Inter-rater reliability for the items in the interview
was high (kappa coef®cient or intra-class correlation at .75 or higher). This level of agreement was
achieved for over 80% of the interview items. Keywords: Diagnosis, diagnostic interview, autism
spectrum, reliability.
The purpose of this paper is to describe the devel- which is by de®nition the fundamental feature of
opment of the Diagnostic Interview for Social and autistic disorders, none of the items on which the
Communication Disorders (DISCO) and to report its diagnoses are based is found exclusively in individ-
inter-rater reliability. The DISCO is a semi-struc- uals with autistic spectrum disorders. Each can oc-
tured interview schedule designed to collect infor- cur in other developmental conditions and can be
mation on development and behaviour. Its purpose seen even in typical development, especially in the
is to assist clinicians in the diagnosis, differential early years (Leekam, 1996; Wing, 1969, 1971).
diagnosis and management of autistic spectrum and The various sets of diagnostic criteria that have
other developmental disorders affecting social inter- been published, including ICD-10 and DSM-IV,
action and communication. rely on a `categorical' rather than a `dimensional'
Autistic spectrum disorders include a wide range principle. That is, diagnosis and differential diag-
of conditions having in common a triad of impair- nosis depend upon identifying a certain number
ments of social interaction, communication and im- and degree of severity of untypical developmental
agination. This triad is associated with a narrow, and behavioural features. The items included
repetitive pattern of activities and interests (Wing, in the different systems were originally based on
1988; Wing & Gould, 1979). The spectrum includes, descriptions given by the clinicians who ®rst con-
but is wider than, the sub-groups in the category of ceptualised that a particular cluster of features
`pervasive developmental disorders' as de®ned in the could be regarded as a syndrome (for example,
tenth edition of the International Classi®cation of Kanner, 1943; Asperger, 1944, 1991). These descrip-
Diseases (ICD-10 research criteria) (World Health tions have subsequently been modi®ed through the
Organisation, 1993) and in the fourth edition of the clinical experience of workers in the ®eld. None of
American Diagnostic and Statistical Manual (DSM- the suggested `syndromes' has been independently
IV) (American Psychiatric Association, 1994). The validated in terms of aetiology, neuropathology or
relationship between autistic spectrum disorders underlying psychological dysfunction. Attempts to
and pervasive developmental disorders and the rea- validate, for example, through ®eld trials such as
sons for preferring the former term are presented in the one reported by Volkmar, Cicchetti, Bregman,
detail by Wing (1997). and Cohen (1992) are inherently circular because
Autistic spectrum disorders can be identi®ed they are based on clinical judgements and not
only on the basis of patterns of developmental and upon any independent criteria.
behavioural features. However, the sets of features Although the internationally accepted standard-
that in theory de®ne sub-groups within the spectrum ised diagnostic and classi®cation systems are useful
in practice overlap with each other. Furthermore, for research studies requiring narrowly de®ned
with the exception of autistic social impairment, groups of participants, clinical work aimed at
Ó Association for Child Psychology and Psychiatry, 2002.
Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA
308 Lorna Wing et al.
helping individuals demands a different, `bottom-up' according to the theories of the nature of autism
approach. Most clinicians working with children or held by their respective authors. None of those
adults see a wide range of different conditions published was suitable for the Camberwell study
among which autistic disorders may form only a because none covered a range of developmental
small proportion. For this type of clinical work the domains as well as behaviour. Moreover, none of
aim is to obtain a detailed developmental history them was organised in a systematic, domain-
from infancy onwards, covering the acquisition of speci®c fashion. All of them were designed to
practical and social skills and patterns of behaviour identify typical autism rather than to record clinical
emerging over the years. The features that are likely pictures in children with varying numbers of aut-
to occur in autistic spectrum disorders can then be istic features.
seen in the full developmental context and their For the foregoing reasons the Handicaps, Beha-
signi®cance more accurately evaluated. Compiling viour and Skills schedule (HBS) was developed for
this type of life history highlights the inter-relation- the Camberwell study (Wing & Gould, 1978, 1979).
ship of developmental level and overt behaviour and This was a semi-structured interview schedule that
how these affect each other in different ways at dif- covered developmental skills including social in-
ferent stages of life (Eisenmajer et al., 1998). Lord teraction, verbal and non-verbal communication as
(1997), in her detailed review of diagnostic instru- distinct from formal language, and imagination. It
ments for autistic spectrum disorders, pointed out also included a wide range of behaviour items, es-
the need for instruments that address diagnoses pecially those found in autistic disorders. The re-
beyond typical autistic spectrum disorder. The sults of this study led to the formulation of the
DISCO attempts to achieve this goal. hypothesis of a spectrum of autistic disorders,
which included Kanner's (1943) classic autism,
Asperger syndrome (Asperger, 1944, 1991; Ehlers
The origins of the DISCO
& Gillberg, 1993; Wing, 1981), the group of `loners'
The earliest origins of the DISCO are to be found in a described by Wolff (1995), and others with mixtures
study comparing children with autism with children of features from all these groups (Wing, 1996). It
with other disabilities or with typical development was suggested that disorders in the spectrum all
(Wing, 1969, 1971; Wing & Wing, 1971). The had in common a triad of impairments of social
`Childhood Behaviour Schedule', a questionnaire interaction, communication and imagination, as-
sent to parents by post, was designed for this study. sociated with a narrow, repetitive pattern of activ-
It elicited information concerning the social, lan- ities. The elements of the triad and the repetitive
guage, imagination and motor impairments and the behaviour could all be manifested in a wide range
odd responses to sensory input and stereotyped of different ways.
behaviour found in autism, which are covered in When the Elliot House Centre for Social and
much more detail in the DISCO. Communication Disorders was set up by the
In the early 1970s Wing and Gould (1978, 1979) National Autistic Society (UK) in 1990, a schedule
carried out a study of autistic disorders among for clinical assessment as well as research was re-
children in the former London Borough of Camber- quired. Both children and adults were referred to
well. Lotter (1966, 1967) had previously completed a Elliot House and the clinical pictures seen included
study of the prevalence of typical autism as des- the whole autistic spectrum and other develop-
cribed by Kanner (1943). He looked speci®cally for mental disorders on the borderlines of the spec-
children showing the narrow criteria for this condi- trum. Individuals with any level of ability from
tion suggested by Kanner and Eisenberg (1956). In profound learning disability to superior cognitive
contrast, the aim of the Camberwell study was to skill were seen. Some who were referred had asso-
identify children with any of the features of autism ciated physical or psychiatric conditions or other
and to see if any patterns could be discovered among developmental disorders such as dyslexia and dys-
the clinical pictures thus found. Lotter's study was praxia. The schedule used had to be relevant for all
`top down', whereas the Camberwell study was `bot- these variations in the clinical pictures. The HBS
tom up'. schedule had been designed for use with children. It
In order to study the range of developmental and recorded information concerning current levels of
behavioural patterns in the children in the Camb- developmental skills and current behaviour. It
erwell study, an interview schedule was required did not have a historical perspective. For these
that covered developmental skills and impairments reasons it was not suitable for use in the Elliot
as well as behaviour. At the time a number of House centre.
schedules purporting to diagnose autism had been Other instruments that had been published after
published. Some were checklists completed by in- the Camberwell study was completed were consid-
formants and others were behaviour observation ered. The Autism Behaviour Checklist (ABC) de-
scales. Freeman and Ritvo (1981) published a designed by Krug, Alrick, and Almond (1980) was
tailed and critical review of such diagnostic sys- widely used. However, it was brief, it required yes or
tems. They pointed out that these systems varied no answers and was designed to identify autism
Diagnostic Interview for Social and Communication Disorders 309
only. The Childhood Autism Rating Scale (CARS)

Development of the DISCO
developed by Schopler, Reichler, and Renner (1986)
was completed by observation of a few crucial as- The choice of items for the DISCO was guided by
pects of behaviour. It was mainly concerned with the the following considerations. First, items relevant
diagnosis of autism and was intended for use with for the basic diagnostic criteria for autistic spec-
children. trum disorders, that is, items concerning social
The Autism Diagnostic Interview (ADI) designed by interaction, communication, imagination and a
Rutter and his colleagues (Le Couteur et al., 1989), narrow repetitive pattern of activities, were inclu-
later revised (ADI-R) (Lord, Rutter, & Le Couteur, ded. Experience with children and adults referred
1994) was the most obvious possibility. This is a to Elliot House showed the wide variation in the
semi-structured parent interview covering in sys- ways in which these impairments could be mani-
tematic fashion developmental and behavioural fested (Wing & Gould, 1979; Wing, 1996). During
items relevant for the diagnosis of autism, which is the course of the DISCO development the relevant
coded for computer analysis. It can be used with sections were expanded to ensure that these vari-
adults as well as children. As noted by Lord et al. ations were covered.
(1994), the original ADI was designed for research Second, when using the HBS schedule and al-
purposes but came to be used also for clinical work, lowing parents to talk around the questions asked,
which was one of the reasons why it was revised. It it had become clear that with children and adults
provides a diagnostic algorithm for the ICD-10 and in the autistic spectrum there was often a marked
DSM-IV de®nition of autism. The aim at Elliot House discrepancy between the apparent level of develop-
was to develop an instrument that would enable ment in different practical skills and the way those
diagnoses to be made of autistic spectrum disorders skills were used in everyday life. For example,
but, more importantly, would help to answer two many individuals preferred their parents to con-
questions. First, `What is this person's pattern of tinue to wash and dress them even when they had
development and behaviour?' and second, `What are the necessary skills and were well past the age
this person's needs?' A new semi-structured inter- when they should have become independent. Items
view schedule, the Diagnostic Interview for Social were included in order to record the level of devel-
and Communication Disorders (DISCO), was development in different domains. The hierarchies of
oped based on the HBS but more detailed and wide- developmental levels in each domain were obtained
ranging, covering past history as well as the present from standard publications on child development.
clinical picture and suitable for all ages and levels of The main source was the Vineland Adaptive Beha-
ability. viour Scales (Sparrow, Balla, & Cichetti, 1984).
The DISCO differs from the ADI in various ways Other sources supplementing the information from
of which the following are the most important. The the Vineland were Cooper, Moodley, and Reynell
ADI was designed primarily as a diagnostic in- (1978), Egan, Illingworth, and MacKeith (1969),
strument whereas the DISCO was designed to ob- Grif®ths (1967), and Sheridan (1973, 1977). The
tain systematically information needed to compile average ages at which a typically developing child
an individual's developmental history from birth would be expected to achieve different steps in de-
and a description of their current pattern of beha- velopment were also obtained from these publica-
viour. Both the ADI and the DISCO record untypi- tions. They were noted in the DISCO beside each
cal behaviour in the past and currently. For many step in each developmental hierarchy. These items
items the ADI speci®es that the coding of past be- were included in the DISCO to help in assessment
haviour refers to that seen at 4±5 years of age. The of needs and also in order to facilitate the rating of
DISCO records the age of onset of deviant devel- the untypical behaviour items related to develop-
opment but for each individual item of untypical mental domains.
behaviour it rates the worst manifestation at Third, clinical experience showed the frequency of
whatever age it occurred. As can be seen by com- many other untypical behaviours associated with the
paring the lists of items in Le Couteur et al. (1989) autistic spectrum but not speci®cally related to de-
and Lord et al. (1994) with those in the Appendix to velopmental domains or diagnostic criteria. These
the present paper, the ADI includes some develop- included odd responses to sensory stimuli, emotional
mental items but the DISCO covers a wider range of disturbance, problems of attention and level of ac-
developmental domains. These include aspects of tivity, `challenging' behaviour and, in adolescents
self-care, independence, visuo-spatial and other and adults, catatonic phenomena (Wing & Shah,
skills. It also includes more detail on untypical 2000), psychiatric disorders, dif®culties associated
responses to sensory stimuli, motor stereotypies, with sexual behaviour and forensic problems. It was
repetitive routines and emotional disturbance, and decided to include sections on these subjects because
has sections on catatonia, psychiatric disorders, of the relevance for understanding the problems ex-
dif®culties related to sexual behaviour and forensic perienced by families and other carers and assess-
problems. ment of the needs of the individuals concerned.
should proceed. Informants are encouraged to des-

Uses of the DISCO
cribe examples of behaviour or to relate illustrative
The DISCO was designed to obtain, systematically, anecdotes.
information needed to compile an individual's clin- As pointed out by Cox and Rutter (1985), the re-
ical history from birth and a description of their call of the timing of events is much less reliable than
current clinical picture. This information can be the memory of their occurrence. Angold, Erkanli,
used to make a clinical diagnosis of a condition Costello, and Rutter (1996) found that, with symp-
within the autistic spectrum, a developmental dis- toms that had lasted a year or more, the recall of the
order on the borderlines of the spectrum, a psychi- date of onset was usually uncertain. Dating aspects
atric disorder or any combination of these of behaviour that occurred long in the past is likely
conditions. It can be used as a guide for recom- to be even more inaccurate. The DISCO does ask for
mendations concerning education, occupation, leis- details of past as well as current behaviour but,
ure and care. For example, for a child of school age, apart from a few items concerning developmental
information concerning discrepancies in develop- skills and any setbacks, the ages when they oc-
mental levels reached in different domains would curred are not coded (see Appendix). The instruc-
provide a guide to the starting points for teaching in tions to the interviewer are to code behaviour that
different areas. the informant remembers easily and clearly. If the
The schedule can also be used for research. For informant has to search their memory and remains
this purpose, provisional algorithms have been writ- uncertain the item should be coded as absent (for
ten. These include three of the sets of standard in- untypical behaviours) or not known (for delays in
ternational diagnostic criteria for pervasive the development of a particular skill). Parents are
developmental disorders and their sub-groups; DSM- encouraged to bring any records they may have of
III-R (American Psychiatric Association, 1987), ICD- their child's development and any photographs and
10 (World Health Organisation, 1993) and DSM-IV videos of their early years, which are often most
(American Psychiatric Association, 1994). In addi- helpful in establishing dates of developmental mile-
tion, algorithms have also been written for three stones. Photos or videos may also show odd facial
diagnostic categories based on Kanner's criteria for expressions or postures, particularly of the arms
early infantile autism (Kanner & Eisenberg, 1956), and hands.
Wing and Gould's (1979) criteria for autistic spec- The collection of the information required for the
trum disorder and Gillberg's criteria for Asperger DISCO should not be con®ned to the interview
syndrome (Ehlers & Gillberg, 1993). A comparison alone. Ratings should be checked and if necessary
between the DISCO 9 algorithms for the ICD-10 cri- changed in the light of information from psycholo-
teria for Asperger syndrome and for Gillberg's criteria gical assessment, structured and unstructured ob-
for this syndrome (Ehlers & Gillberg, 1993) has been servations of the child or adult concerned,
published (Leekam, Libby, Wing, Gould, & Gillberg, interviews with other informants such as teachers
2000). The accompanying paper (Leekam, Libby, and any available case records. The reasons for any
Wing, Gould, & Taylor, 2002) describes a study of the marked discrepancies should be investigated as far
DISCO algorithms for ICD-10 childhood autism and as possible. At Elliot House, with the family's per-
the Wing and Gould autistic spectrum disorder. mission, professionals working with the child or
adult concerned are encouraged to take part in the
assessment. This has been found to be valuable not
Administration of the DISCO
only for obtaining the DISCO information but also
In their examination of techniques of interviewing for planning and implementing services for the
parents about their children, Cox and Rutter (1985) person concerned.
quoted experimental evidence to show that system- More details concerning administration of the
atic questioning obtained fuller and more detailed DISCO are given in the section on methods of rating,
information than free reporting of children's symp- below.
toms. The DISCO schedule is investigator based;
that is, the task of the interviewer is to elicit enough
information from the informant to make a judgement
Study of inter-rater reliability
as to the most appropriate rating for each item
(Wing, Cooper, & Sartorius, 1974). Instructions for Inter-rater reliability for the coding of DISCO items
the interviewers and suggestions for introductory was tested in the study reported below. This was
questions are provided for each item but the wording carried out together with a study of the ability of the
is not ®xed. The questioning has to be adapted in the DISCO to discriminate between autistic disorders
light of the level of functioning of the child or adult and other developmental disorders, reported in the
concerned gained from prior information or during accompanying paper (Leekam et al., 2002).
the course of the interview, and the cultural back- The version of the DISCO used in these studies
ground of the informant. Following the introductory was DISCO 9. The results were used to indicate the
questions the replies determine how the questioning changes needed to improve the reliability of the
ratings in order to produce DISCO 10, in current full achievement. Each step was numbered in serial
use. order, with the average age of achievement in typical
development noted beside the number. The number of
the step currently achieved by the individual concerned
Method was entered on the coding sheet and used for the
The contents of the DISCO. All the items in DISCO 9 computer analysis.
are listed in the Appendix. For the statistical analysis Items concerning developmental delay (Type b) were
presented in this paper, the items were grouped under coded numerically on a three-point scale marked as
four main headings (see Tables 3±6) delay, minor delay or no problem. The exceptions to this
1 Infancy (the ®rst year). were walking, toilet training, saying meaningful words,
2 Age of recognition of deviant development and set- and meaningful phrases, for which the actual ages of
achievement were recorded if known. For certain items
backs in development.
used in the ICD-10 algorithm for Asperger's syndrome,
3 Developmental skills. Fifteen different developmental
the rating concerned whether the individual had ac-
domains were covered. The items were of three kinds: quired the relevant skill by three years of age. For older
Type a) concerned the current level of function in children and adults the `not known' code may have to be
each domain used. Any developmental setbacks were also recorded
Type b) concerned delay in reaching milestones rele- as ages when they occurred.
vant for the domain The items dealing with untypical behaviour (Type c),
Type c) rated untypical behaviour in relation to the including those directly related to developmental do-
domain, such as unwillingness to use a skill despite mains but excluding those observed in the ®rst year of
evidence of its presence. (`Domestic skills' was the life, were scored twice (see Tables 5 and 6 and the
only domain for which no untypical behaviour was Appendix). The ®rst rating coded whether the behaviour
described had ever been present and its most severe
included.)
manifestation, and the second coded whether it was
Type a) and Type b) items are grouped together in
currently present. The same three-point numerical code
sections under the heading `Developmental' as shown was used for each rating. This indicated whether the
in Tables 3 and 4. Type c) items are also shown untypical behaviour, at the time it was present (èver' or
in sections under the heading `Developmental' in `currently'), was severe, minor or not present. The
Tables 5 and 6. severe rating was used when a behaviour was observed
4 Untypical behaviour not directly related to a speci®c every day; or if it occurred whenever carers' strategies
developmental domain. The items of this kind used were not in use (e.g., self-injury if no protective clothing
in the study were grouped into 11 sections under the was worn); or if it occurred whenever the opportunity
heading Òther' as shown in Tables 5 and 6. Quality arose (e.g., inappropriate reaction to visitors). For some
of social interaction was classi®ed as aloofness, behaviours that occurred less frequently but were
severe when they did occur, such as major physical
passivity, active but odd approaches to other people,
aggression, the severe coding was also used. The minor
or social interaction appropriate for overall level
coding was used if the behaviour occurred sometimes.
of development (Wing & Gould, 1979). In clinical For a few items that did not ®t these general rules,
use the rating of this item should be based on the speci®c instructions were included in the DISCO
interviewer's judgement, using information derived manual.
from the rest of the interview, observation of the For some of the items there was the possibility that
child or adult concerned, other informants and case they would not be applicable. For example, if someone
records. had no speech then the items on untypical speech,
The sections concerning adult psychiatric illnesses, such as reversal of pronouns, could not be rated. If a
dif®culties related to sexual behaviour and forensic child was too young to have reached the level of
problems were not used in this study because they were development necessary for a particular skill (e.g., tying
unlikely to be relevant for children in the age ranges shoe laces, using money, as indicated by the average
participating in the study. Catatonic features had been ages for typically developing children shown with the
noted in a number of adolescents and adults seen at item on current developmental level), the rating of
Elliot House (Wing & Shah, 2000). These were much delay in development would not be applicable. If a
less common in children so it was decided to rate only child was too young to understand a social rule (e.g.,
the three items from the section on catatonia that had not making embarrassing remarks in public), this
been observed in more than one child previously seen at would be rated as not applicable in the section on
Elliot House. maladaptive behaviour. For these items, ages by
which the untypical behaviour would be considered
Methods of rating. Each item was rated using a inappropriate were suggested in the instructions to
numerical code. There were three rating scales, one the rater. However, the age rule could not be rigid
for each type of developmental item (Types a, b and c because the rater had to take into account the social
de®ned above). As described above, for the ratings of culture in which the person concerned was living.
current levels of development (Type a), recognisable Another possibility was that the information needed
steps in development of the relevant skill were arranged could not be obtained. Numerical codes were assigned
in chronological order according to the average ages for `not applicable' and `not known' and were available
when they are expected to be achieved, up to the level of for all items.
Some problems were encountered in incorporating Interviewers. Two members of the team based at the
diagnostic criteria from ICD-10 into the DISCO be- University of Kent (SRL and SJL) conducted and coded
cause ICD-10 did not include speci®c instructions or the interviews. They were non-clinical researchers who
guidelines. For example, it would be dif®cult to decide had undergone three months' training on how to use
how to rate the age of onset of meaningful speech the DISCO. Training included coding at least ®ve
(single words and combining words) if a child started diagnostic interviews with one or other of the clinicians
to speak and then stopped but began to speak again who designed the DISCO (LW and JG). In addition,
later. In such cases, arbitrary rules to deal with each before the study, each researcher independently con-
speci®c instance had to be included in the DISCO. (In ducted at least four interviews with parents, which were
the example given, it was decided to enter the age recorded and rated by the clinicians. High levels of
when the child ®rst began to speak.) agreement (86% and 85% of items respectively) were
reached between each coder and clinician.
Participants. The parents of 82 children, 50 school-
age children aged 80 to 140 months and 32 pre-school Preparations for the interviews. Another member of
children aged 34 to 67 months, participated in this the team, based at the Centre for Social and Commu-
study. There were four participant groups; children nication Disorders, Bromley (ML), made pre-interview
diagnosed as having autistic spectrum disorder (18 low contact with the parents. Because of the discrimin-
functioning and 18 high functioning, total 36 partici- ation aspect of the study, he asked them not to reveal
pants), learning disability (17 participants), or speci®c to the interviewers their child's diagnosis, the school
language disorder (14 participants), and a group with they attended or any other information that might
typical development (15 participants). The children with have given away the diagnosis. One interviewer (SJL)
disabilities were recruited from clinics and special was blind to the diagnoses of all the children and knew
schools in London, Kent and Sussex. They had been only the parents' names and addresses before each
diagnosed by clinicians working in these areas and interview began. The other (SRL) was blind to the
none had been assessed or diagnosed by members of diagnoses of the 50 school-age children. The pre-
the research team. The typically developing children school children were simultaneously involved in
attended schools and nurseries in the local area. The another study organised by SRL. She met them and
parents of all the children were given a full explanation their families for about 15 minutes only, to greet them
of the aims and methods of the study and had access, before they were tested by other researchers. She was
on request, to feedback of the results. aware of these children's diagnostic groups for the pur-
The characteristics of the children and results of pose of the other study but had no other information
psychometric tests are shown in Table 1. For the inter- about their diagnostic pro®les. Two parents of school-
rater reliability study, information across all the groups age children and one parent of a pre-school child also
was combined. Further details concerning the children accidentally revealed their child's diagnosis during the
and the procedure for obtaining psychometric informa- interview.
tion are given in the accompanying paper (Leekam Most of the interviews for the school-age children
et al., 2002). The tests used are shown in Table 1. were conducted in the parents' homes, while the
Table 1 Characteristics of the children
Autistic low Learning Autistic high Language Typically

functioning disabled functioning disorder developing
A Characteristics of school-aged children

Chronological age* 106 (80±133) 112.9 (86±140) 107.5 (80±131) 107.9 (80±136) 107.6 (83±135)
(months) N = 11 N=9 N = 11 N=9 N = 10
Male/Female ratio 11:0 5:4 9:2 5:4 6:4
Non-verbal IQ* 54.3 (12±80) 50.0 (14±73) 92.4 (64±140) 95.6 (77±117) 107.6 (92±121)
(Leiter or Bayley) N = 11 N=8 N = 11 N=8 N = 10
Language age* 28.8(10±75) 51.2 (12±87) 89.9 (52±186) 95.4 (62±138) 119.1 (78±186)
(Reynell, WOLD or McArthur) N = 10 N=8 N = 11 N=9 N = 10
(Expression ± months)
Language age* 27.5(8±81) 44.5 (8±87) 84.4 (42±168) 81.4 (24±120) 115.8 (290±156)
(Comprehension ± months) N = 10 N=8 N = 11 N=9 N = 10
B Characteristics of pre-school children
Chronological 49.9(34±58) 51.2 (40±62) 55.4 (35±67) 55.2 (49±62) 53.0 (51±56)
Age* (months) N=7 N=8 N=7 N=5 N=5
Male: Female ratio 6:1 4:4 6:1 4:1 3:2
Non-verbal IQ* 44.0(21±95) 41.7 (20±78) 112.0 (85±129) 86.8 (58±116) 110.6 (81±138)
(Leiter or Bayley) N=7 N =9 N=7 N=5 N=5
Language age* 17.8 (8±39) 22.5 (8±34) 32.4 (12±57) 45.6 (33±67) 62.0 (42±77)
(Reynell, WOLD or McArthur) N=7 N=8 N=7 N=5 N=5
(Expression ± months)
Language age* 12.8 (8±19) 15.3 (9±30) 31.4 (12±61) 35.2 (17±56) 55.6 (33±83)
(Comprehension ± months) N=7 N=8 N=7 N=5 N=5
* For each of these variables, the mean score, the range of scores and the number of participants receiving the test are shown.
interviews for the pre-school group were conducted at applied particularly to the items concerned with
the University of Kent at Canterbury. setback in aspects of development and those related
to special skills sometimes seen in autism, such as the
Design and procedure. Both interviewers were pre- ability to compose music (true of one child in this
sent throughout each interview. The design was coun- study). Items for which three or fewer children were
terbalanced so that half of the interviews within each of rated as having the relevant pattern of development,
the three age bands were conducted by each of the two behaviour or skill were not considered suitable for
researchers. At each interview, one researcher conduc- statistical analysis (see Appendix.).
ted the interview and coded the parent's responses Table 2 shows that 42 ratings were not applicable or
while the other listened and coded the responses, but not analysed for the school-age children and 96 for the
remained silent during the course of the interview. The pre-school group When the `not applicable' or `not
researchers each completed the DISCO coding sheets analysed' ratings were excluded, for the school-age
without conferring and, after each interview, sent the children there were 116 ratings for developmental items
coding sheets to the Centre at Bromley for ML to enter and 350 for the items concerning untypical behaviour.
the data. The numbers for the pre-school children were 100 and
The interview took approximately three hours to 312 respectively (see Table 2).
complete. With the exception of the pre-school children,
mentioned above, who were seen by SRL before the Statistical analysis. The levels of agreement between
interview, the interviewers did not see the children the interviewers were tested by using unweighted
before or during the interview. Because one researcher Cohen's kappa for items with two or three codes. For
was blind to the diagnoses of only the older children, items with four or more codes, intra-class correlations
tests of inter-rater reliability were conducted separately were calculated. Agreement was considered to be high if
for the school-age and the pre-school children. Cohen's kappa or the intra-class correlation was equal
For the sake of clarity, each of the skills or behaviours to or greater than .75.
rated will be referred to as ìtems'. The codes for each of
the two age groups, and for the `current' and èver'
divisions of the untypical behaviours will be referred to Results
as `ratings'.
As shown in the Appendix, there were 319 items There were two stages to the analysis of the items.
included in the interview for the study. There were 130 First, the inter-rater agreement on whether an item
items concerned with developmental skills and 189 for was considered to be not applicable or not known
untypical behaviour. As described above, the latter for a particular child was tested. As noted above,
were rated twice, giving a total of 378 ratings for some ratings were excluded from the analysis be-
untypical behaviour and a grand total of 508 ratings. cause the great majority of the children had `not
However, as already explained, some ratings were not
applicable' codes. However, for some items that
applicable for all or most children in one or both age
were included, there were some children for whom
groups and/or for one or both of the current and ever
ratings. When there were less than 12 children for the `not applicable' or `not known' codes had to
whom a rating was applicable, that rating was exclu- be used. These ratings were examined separately
ded from the analysis. There were also some items for because the `not applicable and `not known' codes
which only a very small number of children were rated were conceptually different from the positive
as having the relevant untypical behaviour. This ratings of a child's behaviour. Sometimes a
Table 2 Number of DISCO ratings and level of inter-rater reliability
School-age children Pre-school children
Kappa or percentage Infancy, age of Untypical Infancy, age of Untypical

agreement recognition and behaviour ratings recognition and behaviour ratings
developmental ratings developmental ratings
No. (%) No. (%) No. (%) No. (%)
High Kappa or intra-class 105 (90.5) 292 (83.4) 89 (89.0) 261 (83.7)
correlation => .75
K or R <.75
0.70±0.74 6 (5.2) 29 (8.3) 3 (3.0) 19 (6.1)
0.60±0.69 3 (2.6) 22 (6.3) 5 (5.0) 21 (6.7)
0.50±0.59 2 (1.7) 3 (0.9) 1 (1.0) 3 (1.0)
0.40±0.49 ± 4 (1.1) 1 (1.0) 4 (1.3)
0.30±0.39 ± ± ± 3 (1.0)
0.20±0.29 ± ± 1 (1.0) 1 (0.3)
TOTAL 116 (100) 350 (100) 100 (100) 312 (100)
Not applicable 14 28 30 66
Grand Total 130 378 130 378
judgement had to be made whether the `not ap- Table 3 School-age children. Number of ratings with high
plicable' or, in some cases, the `not known' code inter-rater agreement for infancy, age of recognition, and
developmental skills
should be used. For example, if a child had only a
few single words the interviewer had to decide how Total no. No. of items with
the item on echolalia should be rated. The DISCO DISCO sections* of items high agreement #
instructions note that if there is insuf®cient speech
INFANCY 25 21
the item on echolalia should be coded as `not ap-
AGE OF RECOGNITION 2 2
plicable'. Since it is not possible in practice to de- DEVELOPMENTAL:
®ne precisely `suf®cient speech', the interviewers Gross motor skills 7 7
had to make a judgement. From among the ratings Self-care: Toilet training 5 5
that were analysed there were `not applicable' and/ Feeding 6 6
Dressing 6 6
or `not known' codes used for some children for
Personal hygiene 3 3
167 ratings for the school-age group and 135 for Domestic skills 4 4
the pre-school group. The inter-rater agreement for Independence 3 3
non-applicability was high (as de®ned above) for Communication:
over 85% of the relevant items for both the school- Receptive 5 5
Expressive 4 4
age and the pre-school children.
Non-verbal 8 7
The second stage of the analysis examined the Social interaction 3 2
absolute inter-rater agreement for each item calcu- Imitation 3 3
lated from the raw ratings that were deemed applic- Imagination 8 8
able by both raters. The Appendix shows the results Skills: Visuo-spatial 9 9
Other 15 10
of the 878 separate calculations of levels of agree-
ment. Table 2 summarizes these results, giving the TOTAL 116 (100%) 105 (83.00%)
numbers of ratings for which agreement was high * Items in these sections rated once only.
(k or r >.75) for school-age and pre-school children # Kappa or intra-class correlation => .75.
for infancy, age of recognition and developmental
skills and for untypical behaviour. The table also
shows the range of kappa or intra-class correlation Table 4 Pre-school children. Number of ratings with high
inter-rater agreement for infancy, age of recognition, and
values for the items for which agreement was not developmental skills
high as de®ned. Agreement was high for 90.5% of
developmental items for the school-age children and Total no. No. of items with
83.4% for the untypical behaviour items. For the DISCO sections* of items high agreement#
pre-school children the levels were 89.0% and 83.7% INFANCY 25 22
respectively. AGE OF RECOGNITION 2 2
Table 3 gives the results for the school-age DEVELOPMENTAL:
children for the items concerning infancy, age of Gross motor skills 6 6
recognition and the developmental domains. Ag- Self-care: Toilet training 5 5
Feeding 5 4
reement was high as de®ned for all items for 13 of Dressing 4 4
the 17 developmental areas listed. There were only Personal hygiene 2 1
11 items in total (9.5%) with lower levels of agree- Domestic skills 3 3
ment. The area with the largest proportion of items Independence 3 3
with lower agreements covered skills such as read- Communication:
Receptive 5 5
ing, writing, number work, drawing and so on. The Expressive 4 4
lowest level of kappa for the 5 items concerned was Non-verbal 8 7
.67 and three were between .70 and .74. One of Social interaction 3 1
the 3 items concerned with development of social Imitation 3 2
interaction, namely, lack of selective social attach- Imagination 6 4
Skills: Visuo-spatial 9 9
ment, had a lower level of inter-rater agreement Other 7 7
(k .54).
TOTAL 100 (100%) 89 (89.0%)
Table 4 shows the level of agreement for pre-
school children for items concerned with infancy, * Items in these sections rated once only.
age of onset of developmental problems, and # Kappa or intra-class correlation = > .75.
developmental skills. Agreement was high for 89%
of the items. There were 11 items (11%) with lev- Under the heading of ìmagination', the items on
els of agreement below .75. For social interaction, `delay in developing pretend play alone' and `delay in
as with the school-age children, there was a asking questions about the world' were below the
problem with `lack of selective social attachment' high level of agreement (k .69 for both).
(k .42). The item `does not show preference The untypical behaviour ratings are shown for the
for carers (currently)' also showed low agreement school-age children in Table 5. Combining the èver'
(k .27). and `current' ratings, agreement was high for 292
Table 5 School-age children: Number of items with high inter-rater agreement for untypical behaviour
Total no. of untypical No. of items with high No. of items with high
DISCO sections* behaviour items agreement ± current# agreement ± ever#
DEVELOPMENTAL:
Gross motor skills 4 3 3
Self-care: Toilet training 3 3 3
Feeding 4 2 2
Dressing 2 2 1
Personal hygiene 4 2 4
Independence 0 0 0
Communication:
Receptive 2 2 2
Expressive 9 9 9
Non-verbal 6 6 6
Social interaction:
With adults 20 13 16
With peers 10 6 7
Social play 6 6 6
Imitation 2 0 2
Imagination 3 3 2
Skills: Visuo-spatial 3 2 3
Other 9 8 8
OTHER:
Motor stereotypies 11 10 10
Sensory stimuli:
Proximal 14 13 13
Auditory 3 3 3
Visual 4 4 4
Repetitive routines 18 11 14
Emotions 8 6 7
Activity pattern 6 4 4
Maladaptive behaviour 18 16 17
Sleep pattern 3 3 3
Catatonic features 2 2 2
Quality of social interaction 1 1 1
TOTAL 175 (100%) 140 (80.0%) 152 (86.9%)
* Items in these sections rated twice ± `current' and èver'.

# Kappa or intra-class correlation = >.75.
(83.4%) of the 350 ratings. The èver' ratings had 23 38 ratings were high and, for the latter, 24 (75.0%)
with lower agreement compared with 35 for the out of 32 ratings.
`current' ratings. For the combined ever and current
ratings of untypical social interaction with adults,
29 (72.5%) of the 40 ratings showed high levels of
Discussion
agreement. For interaction with peers, the propor-
tion was 13 (65.0%) out of 20. For repetitive When the DISCO was rated by two independent re-
routines, 25 (69.0%) of 36 ratings showed high ag- searchers, the overall level of inter-rater reliability for
reement. For these aspects of untypical behaviour, the items proved to be high. Agreement at the level of
which are important for diagnosis, the levels of ag- kappa coef®cient or intra-class correlation >.75
reement were lower than the average for all ratings was found for 397 of the total of 466 (85.2%) of all
combined. ratings concerning the school-age children and 350
For the pre-school children (see Table 6), combi- out of 412 (85.0%) of those concerning the pre-
ning the èver' and `current' ratings of untypical school children.
behaviour, there was high agreement for 261 For the developmental items in the school-age
(83.7%) of the 312 ratings. The numbers of ratings children, the area with the largest proportion of
with lower agreement for the pre-school children items with lower agreements covered skills such as
were 23 for the èver' ratings and 28 for the `current' reading, writing, number work and drawing. These
ratings. As for the school-age children, social in- items were not part of any diagnostic algorithm. In
teraction with adults (though not with peers) and any case, in clinical work, objective information
repetitive routines were the sections with fewer can usually be obtained from school records or
ratings than the average with high levels of inter- psychological testing. Of more concern, because of
rater agreement. For the former, 24 (63.2%) out of their relevance for diagnosis, was the low agree-
Table 6 Pre-school children: Number of items with high inter-rater agreement for untypical behaviour
Total no. of untypical No. of items with high No. of items with high
DISCO section* behaviour items agreement ± current# agreement ± ever#
DEVELOPMENTAL:
Gross motor skills 3 2 3
Self-care: Toilet training 3 3 3
Feeding 4 4 4
Dressing 0 0 0
Personal hygiene 2 1 1
Independence N/A N/A N/A
Communication:
Receptive 2 2 2
Expressive 9 8 7
Non-verbal 6 6 6
Social interaction:
With adults 19 11 13
With peers 7 5 6
Social play 3 3 2
Imitation 2 2 2
Imagination 3 2 3
Skills: Visuo-spatial 3 2 3
Other 6 4 4
OTHER:
Motor stereotypies 11 9 10
Sensory stimuli:
Proximal 14 13 12
Auditory 3 3 3
Visual 4 4 4
Repetitive routines 16 11 13
Emotions 8 7 7
Activity pattern 6 6 5
Maladaptive behaviour 16 15 14
Sleep pattern 3 3 3
Catatonic features 2 2 2
Quality of social interaction 1 0 1
TOTAL 156 (100%) 128 (82.1%) 133 (85.3%)
* Items in these sections rated twice ± for current and for ever behaviour.
# Kappa or intra-class correlation = >.75.
ment for the items related to development of at- most of the items concerned. This highlights the
tachment to carers for both school-age and pre- dif®culties inherent in operationalising subtle and
school children. The problem was lack of precision complex social behaviours as compared with other,
in the instructions to the raters. The items were more speci®c aspects of behaviour covered by the
intended to indicate in the `severe' code the aloof DISCO, such as spending long periods staring
indifference to the presence or absence of carers at bright lights and shiny objects. For example, the
characteristic of the child with typical childhood item on àwareness of other's feelings' infringed
autism, at least in the early years. The wording the rule against combining two concepts, by in-
needed to be revised with clear descriptions of the cluding bizarre responses to shows of emotion
manifestations of this behaviour. Under the head- as well as lack of awareness of feelings in other
ing of ìmagination', also crucial for diagnosis, the people.
items on `delay in developing pretend play alone' The other area in which there were fewer items
and `delay in asking questions about the world' with high levels of agreement and which was also
had kappa values of .69 for pre-school children but important for diagnosis was repetitive routines. The
the levels of agreement for school-age children were problem in this case was that dif®cult decisions often
high. had to be made concerning which item best des-
For the untypical behaviours concerned with cribed the routine in question. For example, one
social interaction with adults and, for school-age child was fascinated by electric fans, looked for them
children only, interaction with age peers, the in every house or shop he visited, stared long and
numbers of ratings with high levels of agreement hard at any ceiling fans he saw, demanded electric
were smaller than the average for all ratings com- fans for every birthday and Christmas, and would
bined. Examination of the instructions for the have carried one with him all the time if he could.
interviewers showed that clari®cation of the word- The DISCO rule was to rate speci®c types of un-
ing and provision of examples were required for typical behaviour under one heading only. For this
child, the dilemma was whether to rate under high functioning individuals with social and com-
clinging to objects, collecting objects, or fascination munication problems in subtle forms. Agreement
with speci®c objects. However, from the point of view was also high for the developmental items and for the
of diagnosis, the choice did not matter, because one untypical behaviours related to developmental
marked routine was enough to qualify for this part of domains. These items are not dealt with in the same
the diagnosis of an autistic disorder. As described in detail in other diagnostic schedules known to the
the accompanying paper on diagnostic algorithms authors. They are particularly helpful for assessing
(Leekam et al., 2002), there were high levels of inter- an individual's speci®c problems and needs, what-
rater agreement when social interaction items and ever his or her diagnosis.
the repetitive routine items selected for the ICD-10 The study was limited in its scope. It did not
childhood autism algorithm were collapsed together include adolescents and adults as participants so
according to the algorithm rules. the sections relevant for these age groups were not
The raters had no access to any other sources of examined. There were no tests of agreement of rat-
information concerning the children. One rater only ings made at interviews of the same parents by dif-
had met the pre-school children brie¯y and knew ferent interviewers. Interviews of two different
their clinical diagnoses. Apart from this, the raters informants concerning the same child in different
had no opportunity to meet, observe and interact situations were also not carried out. These last
with the children. More sources of information would be of interest in a clinical context but would
would have made rating easier, especially in rela- present particular problems for testing reliability
tion to social interaction. In these respects, the because of possible environmental effects on chil-
conditions for the research interviews were quite dren's behaviour (Wing & Gould, 1978).
unlike those when using the DISCO in clinical Despite its limitations, the lessons learnt from
work. Despite this, the overall inter-rater agreement the study have been most helpful in indicating
was high. where changes in the schedule were required.
It was possible that the fact that one interviewer However, items of untypical behaviour or special
had met each of the pre-school children for 15 skills that were rarely present in the children in
minutes before the study and knew their clinical this study were retained. This was because, ®rst,
diagnoses, while the other had never seen them and some are seen more often in older and more able
knew nothing about them, may have had some ef- people with autistic disorders and, second, they
fect on the ratings. However, the pattern of inter- are of clinical relevance when they do occur. The
rater agreements was very similar for the two age wider use of the revised schedule (DISCO 10) by
groups. trained professional workers will, it is hoped, pro-
In the parallel study of the ICD-10 childhood vide information that can help to further improve
autism and Wing and Gould autistic spectrum dis- the schedule and give more opportunities for its
order algorithms, no systematic differences were use in clinical research, as suggested in the
found between the two age groups either for the accompanying paper on diagnostic algorithms
ratings used in the algorithms or for the diagnostic (Leekam et al., 2002).
outputs (Leekam et al., 2002).
It is of interest that, in the case of the school-age
Acknowledgements
children, the inter-rater agreements were high for
more of the èver' than for the `current' ratings of This research was carried out while Susan Leekam
untypical behaviour (35 compared with 23), whereas and Sarah Libby were employed at the University of
the numbers were closer for the pre-school children Kent, Canterbury, and Michael Larcombe at the
(28 versus 23). The oldest child in the school-age Centre for Social and Communication Disorders.
group was 11 years 4 months old so the parents' The research was supported by grant SPGS 262
memories of the most severe untypical behaviour from the NHS Executive and by a grant from the
were still relatively clear. Many children with autism Blond Trust. Grateful thanks are due to all the
and other developmental disorders tend to improve parents and children who participated in this study.
in skills and behaviour after they begin school. We would also like to thank the staff of the agencies
Clinical experience with the DISCO interview indi- who helped us to contact the families involved. We
cates that parents are usually very clear about what also thank Colin Taylor, Hannah Farrimond, and
a particular behaviour was like at its worst but tend Chris Ramsden for their expert help with the pre-
to be less clear when describing the degree to which paration of data.
behaviour has subsequently improved. With the pre-
school children, there was less chance of a change
Correspondence to
from èver' to `current' ratings.
For the untypical behaviours, agreement was as Lorna Wing, Centre for Social and Communication
high for the ratings of `minor' as for `marked' prob- Disorders, Elliot House, 113 Masons Hill, Bromley,
lems, suggesting that the DISCO is appropriate for Kent BR2 9HT, UK. Email: [email protected]
Leekam, S.R. (1996). Features of autism in normally

References
developing children. Paper presented at 5th Congress
American Psychiatric Association (1987). Diagnostic Autism±Europe, Barcelona, April.
and statistical manual of mental disorders: Third Leekam, S., Libby, S., Wing, L., Gould, J., & Gillberg, C.
edition ± revised (pp. 33±39). Washington, DC: Ameri- (2000). Comparison of ICD-10 and Gillberg's criteria
can Psychiatric Asociation. for Asperger syndrome. Autism, 4, 11±28.
American Psychiatric Association (1994). Diagnostic Leekam, S., Libby, S., Wing, L., Gould, J., & Taylor, C.
and statistical manual of mental disorders: Fourth (2002). The Diagnostic Interview for Social and Com-
edition (pp. 65±78). Washington DC: American Psy- munication Disorders: Algorithms for ICD-10 child-
chiatric Association. hood autism and Wing and Gould autistic spectrum
Angold, A., Erkanli, A., Costello, E.J., & Rutter, M. disorder. Journal of Child Psychology and Psychiatry,
(1996). Precision, reliability and accuracy in the 43, 327±342.
dating of symptom onsets in child and adolescent Lord, C. (1997). Diagnostic instruments in autistic
psychopathology. Journal of Child and Adolescent spectrum disorders. In D.J. Cohen & F.R. Volkmar
Psychiatry, 37, 657±664. (Eds.), Autism and pervasive developmental disorders
Asperger, H. (1944). Die àutistischen psychopathen' im (2nd edn, pp. 460±483). New York: Wiley.
kindersalter. Archives fur Psychiatrie und Nervenk- Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism
rankheiten, 117, 76±136. diagnostic interview-revised: A revised version of a
Asperger, H. (1991). Àutistic psychopathy' in childhood diagnostic interview for caregivers of individuals
(translated by U. Frith). In U. Frith (Ed.), Autism and with possible pervasive developmental disorders.
Asperger syndrome (pp. 37±92). Cambridge: Cam- Journal of Autism and Developmental Disorders, 24,
bridge University Press. 659±686.
Cooper, J., Moodley, M., & Reynell, J. (1978). Helping Lotter, V. (1966). Epidemiology of autistic conditions in
language development. Arnold: London. young children: I. Prevalence. Social Psychiatry, 1,
Cox, A., & Rutter, M. (1985). Diagnostic appraisal 124±137.
and interviewing. In M. Rutter & L. Hersov (Eds.), Lotter, V. (1967). Epidemiology of autistic conditions
Child and adolescent psychiatry: Modern approa- in young children: II. Some characteristics of
ches (2nd edn, pp. 233±248). Oxford: Blackwell the parents and children. Social Psychiatry, 1, 163±
Scienti®c. 173.
Egan, D., Illingworth, R.S., & MacKeith, R.C. (1969). Schopler, E., Reichler, R.J., & Renner, R. (1986 ). The
Developmental screening 0±5 years. Clinics in devel- Childhood Autism Rating Scale (CARS) for diagnostic
opmental medicine No. 30. London: Heinemann. screening and classi®cation of autism. New York:
Ehlers, S., & Gillberg, C. (1993). The epidemiology Irving.
of Asperger syndrome. A total population study. Sheridan, M.D. (1973). Children's developmental pro-
Journal of Child Psychology and Psychiatry, 34, gress. Windsor: N.F.E.R.
1327±1350. Sheridan, M.D. (1977). Spontaneous play in early
Eisenmajer, R., Prior, M., Leekam, S., Wing, L., Ong, B, childhood. Windsor: N.F.E.R.
Gould, J., & Welham, M. (1998). Delayed language Sparrow, S., Balla, D., & Cicchetti, D. (1984). Vineland
onset as a predictor of clinical symptoms in pervasive Adaptive Behaviour Scales: Circle Pines, Minnesota:
developmental disorders. Journal of Autism and American Guidance Service Inc.
Developmental Disorders, 28, 527±534. Volkmar, F., Cicchetti, D., Bregman, J., & Cohen, D.
Freeman, B.J., & Ritvo, E.R. (1981). The syndrome (1992). Three diagnostic systems for autism: DSM-III,
of autism: A critical review of diagnostic systems, DSM-III-R, & ICD-10. Journal of Autism and Devel-
follow-up studies, and the theoretical background opmental Disorders, 22, 483±492.
of the Behaviour Observation Scale. In J. Gilliam Wing, J., Cooper, J., & Sartorius, N. (1974). Measure-
(Ed.), Autism: Diagnosis, instruction, management ment and classi®cation of psychiatric symptoms.
and research (pp. 17±61). Spring®eld, Illinois: Cambridge: Cambridge University Press.
C.C. Thomas. Wing, L. (1969). The handicaps of autistic children: A
Grif®ths, R. (1967). The abilities of babies. London: comparative study. Journal of Child Psychology and
University of London Press. Psychiatry, 10, 1±40.
Kanner, L. (1943). Autistic disturbances of affective Wing, L. (1971). Perceptual and language develop-
contact. Nervous Child, 2, 217±250. ment in autistic children: A comparative study. In
Kanner, L., & Eisenberg, L. (1956). Early infantile M. Rutter (Ed.), Infantile autism: Concepts, charac-
autism, 1943±1955. American Journal of Orthopsy- teristics and treatment (pp. 173±198). London:
chiatry, 26, 55±65. Churchill.
Krug, D., Arick, J., & Almond, P. (1980). Beha- Wing, L. (1981). Asperger's syndrome: A clinical
viour checklist for identifying severely handicapped account. Psychological Medicine, 11, 115±129.
individuals with high levels of autistic behaviour. Wing, L. (1988). The continuum of autistic character-
Journal of Child Psychology and Psychiatry, 21, 221± istics. In E. Schopler & G. Mesibov (Eds.), Diagnosis
229. and assessment in autism (pp. 91±110). New York:
Le Couteur, A., Rutter, M., Lord, C., Rios, P., Robertson, Plenum.
S., Holdgrafer, M., & McLennan, J. (1989). Autism Wing, L. (1996). The autistic spectrum. London: Con-
Diagnostic Interview: A standardized investigator stable.
instrument. Journal of Autism and Developmental Wing, L. (1997). Syndromes of autism and atypical
Disorders, 19, 363±387. development. In D. Cohen & F. Volkmar (Eds.), Hand-
book of autism and pervasive developmental disorders: Appendix

Second edition (pp. 148±172). New York: Wiley.
Wing, L., & Gould, J. (1978). Systematic recording of All the items from the DISCO used in the study of
behaviours and skills of retarded and psychotic inter-rater agreement are listed in this Appendix.
children. Journal of Autism and Childhood Schizo- Developmental items have one rating for each age
phrenia, 8, 79±97. group. Untypical behaviour items have two ratings,
Wing, L., & Gould, J. (1979). Severe impairments of one under `current' and one under èver', for each age
social interaction and associated abnormalities in group.
children: Epidemiology and classi®cation. Journal of The items under the headings Àlgorithm items'
Autism and Developmental Disorders, 9, 11±29. are those used in any one or more of the algorithms
Wing, L., & Shah, A. (2000). Catatonia in autistic
for ICD-10, DSM-III-R, DSM-IV, Kanner's early
spectrum disorders. British Journal of Psychiatry,
176, 357±362.
infantile autism, Gillberg's criteria for Asperger
Wing, L., & Wing, J.K. (1971). Multiple impairment in syndrome, or Wing and Gould's autistic spectrum
early childhood autism. Journal of Autism and Child- disorder.
hood Schizophrenia, 1, 256±266.
World Health Organisation (1993). The ICD-10 classi®-
cation of mental and behavioural disorders. Diagnos-
tic criteria for research. Geneva: World Health
Organisation.
Wolff, S. (1995). Loners: The life path of unusual
children. London: Routledge.
Manuscript accepted 22 October 2001
Inter-rater agreement for DISCO items
SCHOOL-AGE PRE-SCHOOL
INFANCY
Non-algorithm items
Concern at perinatal condition k = .77 k = .33
Poor sucking k = .77 k = .90
Vomited frequently k = .91 k = .85
Cried excessively k = .62 k = .77
Woke up screaming k = .82 k = .86
Did not sleep well k = .80 k = .93
Unusually good/quiet k = .59 k = .64
Did not demand social attention k = .92 k = .94
Lack of response to cuddling k = .90 k = 1.00
Poor eye contact k = .95 k = .83
Did not anticipate being lifted k = .96 k = 1.00
Lack of reciprocation in lap games k = .97 k = .88
Delay in waving bye-bye k = .85 k = .89
Limited babbling k = .94 k = .95
Limited pre-speech conversation k = .97 k = .94
Lack of response to speech k = .73 k = .85
Lack of intonation in babbling k = 1.00 k = .94
Lack of interest in environment k = .97 k = .74
No sharing of interests k = 1.00 k = 1.00
Did not look when other pointed k = .93 k = .94
Had a special interest k = .84 k = .77
Intensely attached to one parent k = .92 k = .84
Disliked care procedures k = .75 k = .91
Oversensitive to noise k = .80 k = .78
Odd movements of body or limbs k = .70 k = .77
AGE OF ANY SETBACK IN DEVELOPMENT
Algorithm items
Setback in language development 2 children 1 child
Setback in play development 3 children 2 children
Setback in social development r = .95 3 children
Non-algorithm items
Setback in motor development 2 children 1 child
Setback in toilet training 1 child 1 child
Setback in other self-care skills 1 child 1 child
Appendix Table Continued
Setback in range of foods eaten 2 children 1 child

Setback in other adaptive skills none none
Setback in interest and curiosity 2 children 1 child
Age head growth arrested none none
Age of onset of disturbed behaviour r = .84 r = .82
Setback occurred after 24 months 1 child 1 child
Skills lost or not used? 1 child 1 child
Interviewers opinion of onset age 3 children r = .89
Current Ever Current Ever
GROSS MOTOR SKILLS
Algorithm items
Clumsy k = .95 k = .97 k = .90 k = .90
Has immature gait k = .82 k = .86 k = .94 k = .88
Poor coordination in PE and games k = .70 k = .70 NA NA
Non-algorithm items
Ability to walk on level surfaces r = .99 r = .95
Ability to walk up and down stairs r = .88 r = .93
Ability to ride a bicycle or a tricycle r = .98 r = .93
Age sat up independently r = .85 r = .99
Age walked independently r = .99 r = .99
Age rode a tricycle r = .82 r = .99
Age rode a bicycle r = 1.00 NA
Climbs without concern for danger k = .95 k = .91 k = .74 k = .79
SELF-CARE
Self-care: Toilet training
Non-algorithm items
Incontinence during the day r = .93 r = .99
Incontinence at night r = .95 r = .96
Cleans and dresses self after the toilet k = .86 k = .76
Age clean and dry in day r = .90 r = .86
Age clean and dry at night r = .85 r = .91
Resists using pot or lavatory k = .75 k = .80 k = 1.00 k = .93
Smears k = 1.00 k = .89 k = .91 k = .91
Retention of faeces k = .86 k = .95 k = .89 k = .91
Self-care: Feeding
Non-algorithm items
Ability to feed self r = .99 r = .99
Ability to drink from a cup r = .97 r = .88
Delay in eating solid food k = .78 k = .83
Delay in giving up the bottle k = .82 k = .58
Delay in feeding self with spoon and fork k = .88 k = .92
Delay in using knife and fork k = .75 NA
Refuses food which is lumpy or hard to chew k = .64 k = .72 k = .80 k = .83
No interest in food k = .70 k = .74 k = .82 k = .85
Drinks excessive amount of ¯uids k = .93 k = .94 k = .75 k = .86
Refuses to feed self k = 1.00 k = .89 k = .87 k = .87
Self-care: Dressing
Non-algorithm items
Ability to dress self r = .99 r = .97
Ability to do up buttons r = .92 r = .92
Ability to tie laces k = .97 k = .84
Delay in pulling pants down and up k = .90 k = .83
Delay in independent dressing not buttons k = .84 NA
Delay in tying laces k = .92 NA
Unwilling to dress self k = .81 k = .73 NA NA
Excessively slow at dressing k = .96 k = .96 NA NA
Self-care: Personal hygiene
Non-algorithm items
Ability to wash self r = .97 r = .95
Delay in getting hands clean and dry k = .83 k = .72
Delay in bathing and drying self k = .83 NA
Dislikes being washed k = .73 k = .78 k = .75 k = .85
Unwilling to wash self k = .74 k = .77 NA NA
Lack of awareness of need for cleanliness k = .86 k = .86 NA NA

Dislikes sticky or dirty hands k = .80 k = .93 k = .66 k = .68
Self-care: Domestic skills
Non-algorithm items
Helping around the house r = .98 r = .91
Helping with cooking or other creative tasks r = .95 r = .88
Delay in fetching, carrying, taking a message k = .87 k = 1.00
Delay in doing a simple task alone k = .77 NA
Self-care: Independence
Non-algorithm items
Lack of understanding of danger r = .98 r = .95
Amount of supervision required r = .96 r = .93
Can stay at home alone NA NA
Delay in going into the garden alone k = .86 k = 1.00
Delay in going to local shops alone NA NA
Lack of common sense NA NA NA NA
COMMUNICATION
Communication: Receptive
Algorithm items
Level of understanding of language r = .97 r = .96
Limited appreciation of humour r = .94 r = .87
Takes language literally k = .95 k = .97 k = 1.00 k = 1.00
Delay in obeying instructions k = .86 k = 1.00
Non-algorithm items
Lack of understanding of future events k = .83 k = 1.00
Parents concerned about hearing k = .84 k = 1.00
Communication is one-sided k = .95 k = .81 k = .89 k = .76
Communication: Expressive
Algorithm items
Level of expressive language r = .98 r = .98
Age started to use meaningful words r = .97 r = .94
Age started to use phrases r = .78 r = .98
Speech absent or limited & no non-verbal communication NA NA k = 1.00 k= .73
Communication is one-sided k = .86 k= .87 k = 1.00 k= .76
Immediate echolalia k = .87 k= .92 k = .81 k= .87
Delayed echolalia k = .96 k= 1.00 k = .92 k= 1.00
Reversal of pronouns k = .87 k= .91 k = .91 k= .92
Idiosyncratic use of words or phrases k = .84 k= 1.00 k = .46 k= .46
Long winded pedantic speech k = 1.00 k= 1.00 NA NA
Content of speech is irrelevant k = 1.00 k= .92 k = .81 k= .91
Non-algorithm items
Intelligibility of speech r = .98 r = .94
Muddles sequence of words and phrases k = .90 k = 1.00 k = .90 k = .90
Talks to self k = .93 k = .93 k = .93 k = .93
Communication: Non-verbal
Algorithm items
Lack of imperative gestures r = .82 r = .78
Lack of joint referencing pointing k = .87 k = .89
Doesn't nod or shake head k = .95 k = .95
Lack of instrumental gestures k = .89 k = .89
Lack of descriptive gestures r = .87 r = .88
Lack of emotionally expressive gestures k = .77 k = .71
Odd tone of voice k = .84 k = .92 k = .89 k = .82
Uses different voices for no reason k = .88 k = .89 k = .89 k = .79
Facial expression absent or odd k = .76 k = .86 k = .92 k = .86
Non-verbal communication is absent or odd k = .85 k = .81 k = .82 k = .85
Non-algorithm items
Dif®culty understanding gestures r = .92 r = .91
Delay in sharing interests k = .73 k = .77
Good at copying accents, voices and noises k = .88 k = .88 k = .76 k = .76
Does not produce natural smile on request k = .92 k = .77 k = .78 k = .75
SOCIAL INTERACTION
Social interaction: With adults
Algorithm items
Delay in selective social attachment k = .54 k= .42
Odd or no response to physical affection k = .75 k = .76 k= .90 k= .95
Eye contact poor k = .89 k = .90 k= .89 k= .76
Makes brief glances k = .90 k = .86 k= .94 k= .94
Blank unfocussed gaze k = .93 k = .82 k= .86 k= .76
Stares too long and hard k = 1.00 k = 1.00 k= .89 k= .89
Does not greet parents / carers k = .91 k = .91 k= .44 k= .76
Does not greet other people k = .74 k = .81 k= .74 k= .75
Unusual response to visitors k = .74 k = .82 k= .76 k= .76
Uses other people as mechanical aids k = .81 k = .93 k= .89 k= .89
Does not seek comfort when in pain or distress k = .72 k = .76 k= .65 k= .52
Does not seek social / physical comfort k = .68 k = .70 k= .84 k= .86
Does not give comfort to others k = .83 k = .90 k= .73 k= .72
Makes one sided social approaches k = .76 k = .76 k= .94 k= 1.00
Lack of awareness of others' feelings k = .73 k = .71 k= .74 k= .73
Laughs at others' distress k = .87 k = .96 k= .72 k= .66
Unusual response to change caused by injury k = .74 k = .76 k= .32 k= .26
Detached intellectual interest in injury or violence k = .62 k = .62 k= .69 k= .69
Does not share other's happiness k = 1.00 k = 1.00 NA NA
Sharing interests limited or absent k = .88 k = .48 k= .89 k= .85
Non-algorithm items
Does not show preference for carers k = .90 k = .27
Lack of separation anxiety k = .77 k = .89 k = .77 k = .78
Age sharing interests began r = .99 r = .99
Social interaction: With age peers
Algorithm items
No interest in other children k = .88 k = .93 k= .88 k= .90
Does not interacts with peers k = .94 k = .90 k= .89 k= .81
Inappropriate interactions with others k = .80 k = .84 k= .95 k= 1.00
No emotional response to peers k = .94 k = .65 k= .89 k= 1.00
Ignores conventions of peer interaction k = .72 k = .65 k= .74 k= 1.00
Uses other children as mechanical aids k = .86 k = .94 k= .73 k= .69
Lack of friendships with age peers k = .69 k = .89 NA NA
Friendship is not reciprocal k = .73 k = .92 NA NA
Non-algorithm items
Has a preference for older or younger children k = .43 k = .43 k = .83 k = .83
Bullied and teased by age peers k = .80 k = .87 NA NA
Social interaction: Social play
Algorithm items
No reciprocation in simple physical play k = .94 k = .76 k = .86 k = .62
Resists taking turns k = .90 k = .81 k = .89 k = .80
Extreme reaction to losing k = 1.00 k = .95 NA NA
No participation in group / team games k = .81 k = .78 k = .77 k = .77
No participation in organised social activities k = 1.00 k = 1.00 NA NA
Non-algorithm item
Does not play board games k = .92 k = .87 NA NA
IMITATION
Algorithm items
Did not imitate domestic activities in early childhood r = .98 r = .98
Does not imitate actions spontaneously k = .64 k = .79 k = .82 k = .79
Delayed imitation absent or odd k = .71 k = .87 k = .92 k = .92
Non-algorithm items
Delay in waving good-bye k = .90 k = .80
Delay in copying domestic activities k = .88 k = .68
IMAGINATION
Algorithm items
Lack of pretend play r = .98 r = .95
Delay in pretend play alone k = .83 k = .69
Pretend play is repetitive k = .91 k = .95 k = .79 k = 1.00
Non-algorithm items
Doesn't follow stories on TV r = .94 r = .90
Doesn't enjoy listening to stories r = .99 NA
Doesn't ask questions about the world r = .95 r = .95
Delay in imaginative play with peers k = .84 k = .84
Delay in asking questions about the world k = .80 k = .69
Delay in enjoying stories at 5 year old level k = .85 NA
Abnormal reaction to pets k = .86 k = .74 k = .67 k = 1.00
Curiosity limited or absent k = .89 k = .78 k = .90 k = 1.00
SKILLS
Skills: Visuo-spatial skills
Algorithm items
Poor hand-eye co-ordination r = .98 r = .91
Clumsy at ®ne motor tasks k = .88 k = .88 k = .93 k = .87
Non-algorithm items
Use of hands r = .76 r = .96
Visuo-spatial ability e.g., jigsaws r = .98 r = .83
Ability to use scissors k = .89 k = .89
Ability to make 3-D models r = .91 r = .91
Ability to draw r = .99 r = .99
Content of drawings r = .97 r = .89
Delay in completing a 10 piece jigsaw k = .93 k = 1.00
Delay in drawing a recognisable object k = .87 k = 1.00
Unusually dextrous and quick with hands k = .81 k = .82 k = .70 k = 1.00
Reluctant to use hands 1 child 3 childr. none 2 childr.
Reluctant to use pencils or crayons k = .71 k = .78 k = .90 k = 1.00
Skills: Pictures, reading and writing
Non-algorithm items
Ability to recognise scene in picture k = .95 k = 1.00
Reading ability r = .96 r = .99
Writing ability r = .99 r = .99
Delay in identifying objects in pictures k = .81 k = .83
Delay in reading ®rst book k = .76 NA
Delay in writing words without copying k = .74 NA
Does not look at a variety of pictures k = .86 k = .75 k = .80 k = .81
Fascinated with letters or words 3 childr. 3 childr. k = .80 k = 1.00
Not interested in reading k = .90 k = .90 NA NA
Not interested in writing k = .80 k = .76 NA NA
Speci®c reading / spelling impairment k = .72 k = .65 NA NA
Skills: Number, money, dates and time
Non-algorithm items
Numerical ability r = .99 r = .99
Understanding of money k = .84 k = .80
Ability to identify coins k = .84 NA
Knows the days, months and years k = .90 k = 1.00
Ability to tell the time k = .93 NA
Delay in doing simple addition k = .67 NA
Delay in identifying coins k = .69 NA
Delay in knowing the days of the week k = .74 NA
Delay in telling the half hours on the clock k = .70 NA
Skills: Special abilities above general level of development
Non-algorithm items
Special skills: Sings 1 child 1 child 1 child 1 child
Special skills: Plays musical instrument 1 child 1 child none none
Special skills: Composes music 1 child 1 child none none
Special skills: Numerical calculations 1 child 1 child none none
Special skills: Calendar calculations none none none none
Special skills: Visuo-spatial skills k = .87 k = .76 k = .81 k = .81
Special skills: Computers k = .90 k = .95 2 childr. 2 childr.
Special skills: Constructional k = .76 k = .76 2 childr. 2 childr.
Special skills: Drawing 3 childr. 3 childr. none none
Special skills: Maps 1 child 1 child none none
Special skills: Operating equipment k = .78 k = 1.00 none none
Special skills: Memory for verbal material 2 childr. 2 childr. k = .37 k = .37
Special skills: Memory for routes 3 childr. 3 childr. k = .72 k = .72

Special skills: Memory for past events k = .86 k = .86 k = .90 k = .90
MOTOR AND VOCAL STEREOTYPIES
Algorithm items
Jumps up and down with excitement k = .88 k = .89 k = .88 k = .88
Makes unusual movements of hands or arms k = 1.00 k = 1.00 k = .79 k = .83
Midline hand stereotypies k = .93 k = .93 k = .53 k = .63
Spins self around k = .93 k = .93 k = .93 k = .94
Rocks whilst standing up k = .83 k = .83 k = 1.00 k = 1.00
Has complex twisting or rocking movements k = .67 k = .60 k = 1.00 k = 1.00
Non-algorithm items
Rocks sitting down k = .82 k = .87 k = .73 k = .76
Walks on tip toe k = 1.00 k = .91 k = .75 k = .80
Makes sudden jerky movements k = 1.00 k = .89 k = .88 k = .90
Makes shrieks and odd noises k = .84 k = .85 k = .83 k = .83
Makes odd facial grimaces k = .85 k = .91 k = .93 k = .93
RESPONSES TO SENSORY STIMULI
Responses to proximal stimuli
Algorithm items
Self-injury k = .91 k = .93 k = .83 k = .84
Stimulates self without injury k = .75 k = .77 k = .86 k = .72
Smells objects or people k = .89 k = .86 k = 1.00 k = 1.00
Unusual interest in the feel of surfaces k = .84 k = .79 k = .88 k = .94
Aimless and repetitive manipulation of objects k = .51 k = .68 k = .78 k = .80
Non-algorithm items
Smears, urinates to make puddles etc k = .84 k = .96 k = .90 k = 1.00
Places objects in mouth k = .91 k = 1.00 k = .95 k = 1.00
Scratches and taps surfaces k = 1.00 k = 1.00 k = 1.00 k = 1.00
Tears or breaks things in an aimless way k = .91 k = .91 k = .94 k = 1.00
Likes being spun around k = .93 k = .93 k = .84 k = .84
Indifference to pain, heat, cold k = .82 k = .79 k = .62 k = .65
Odd reaction to gentle touch k = .86 k = .86 k = .90 k = 1.00
Odd reaction to a ®rm touch k = .79 k = .80 k = .79 k = .92
Overbreathing k = .76 k = .80 k = 1.00 k = 1.00
Responses to auditory stimuli
Algorithm items
Fascination with sounds k = .87 k = .89 k = .89 k = 1.00
Non-algorithm items
Upset by certain sounds k = .87 k = .77 k = .79 k = 1.00
Has acute hearing k = .95 k = .95 k = .94 k = 1.00
Responses to visual stimuli
Algorithm items
Interested in bright lights and shiny things k = .89 k = .85 k = 1.00 k = 1.00
Interested in watching things spin k = .91 k = .94 k = .89 k = 1.00
Twists hands or objects near eyes k = .91 k = .93 k = 1.00 k = 1.00
Studies the angles of objects k = 1.00 k = 1.00 k = 1.00 k = 1.00
REPETITIVE ROUTINES AND RESISTANCE TO CHANGE
Algorithm items
Clings to objects k = .78 k = .90 k= .65 k= .67
Collects objects k = .79 k = .78 k= .44 k= .44
Fascinated with a speci®c object k = .77 k = .83 k= 1.00 k= 1.00
Arranges objects in patterns k = .86 k = .81 k= .88 k= .89
Interested in the parts of objects k = .60 k = .60 k= .81 k= .92
Complex repetitive activities with objects k = .70 k = .77 k= .79 k= .79
Interested in the abstract properties of objects k = .76 k = .82 k= .79 k= .79
Insists on sameness in the environment k = .60 k = .65 k= .69 k= .76
Insists on perfection k = .87 k = .92 k= .94 k= .94
Has unusual food fads k = .74 k = .73 k= .62 k= .66
Insists on sameness in routines k = .74 k = .79 k= .89 k= .86
Asks repetitive questions k = .90 k = .90 k= .91 k= .91
Talks about a repetitive theme k = .80 k = .77 NA NA
Clings to home k = 1.00 k = 1.00 k= 1.00 k= .91
Acts out role of object, person repetitively k = .60 k = .76 k = 1.00 k = 1.00
Has repetitive activities related to special skill k = .65 k = .56 k = .44 k = 1.00
Collects facts on a speci®c subject k = .82 k = .82 NA NA
Fascinated with TV/ video k = .86 k = .86 k = .85 k = .84
EMOTIONS
Non-algorithm items
Lack of emotional expression k = .79 k = .77 k = .69 k = .69
Frequently unhappy for no reason k = .83 k = .80 k = 1.00 k = .88
Has mood swings k = .89 k = .88 k = .94 k = 1.00
Cries or moans for no reason k = .68 k = .78 k = .91 k = .92
Laughs for no reason k = .89 k = .87 k = .94 k = .95
Puzzled for no reason k = .96 k = .96 k = .89 k = .89
Frequently anxious k = .79 k = .81 k = .86 k = .86
Has an extreme fear of something k = .73 k = .66 k = .79 k = .78
ACTIVITY PATTERN
Algorithm item
Limited pattern of self-chosen activity k = .69 k = .73 k = .89 k = .95
Non-algorithm items
Limited attention span with self- chosen activities k = .79 k = .84 k = .89 k = .95
Limited attention span with activities chosen by others k = .75 k = .81 k = .86 k = .95
Rarely or never sits down k = .42 k = .73 k = .81 k = .74
Continually restless k = .81 k = .89 k = .90 k = .90
Hyperactive k = .83 k = .81 k = .84 k = .75
MALADAPTIVE BEHAVIOUR
Algorithm items
Dif®cult behaviour in public places k = .84 k = .89 k= 1.00 k= .86
Lack of personal modesty k = .89 k = .79 k= .67 k= .67
Doesn't understand psychological barriers k = .92 k = .93 k= 1.00 k= .77
Inappropriate approaches to strangers k = .91 k = .88 k= 1.00 k= 1.00
Makes embarrassing remarks in public k = 1.00 k = 1.00 NA NA
Interrupts conversations k = .80 k = .80 NA NA
Inappropriate response to others' emotions k = .69 k = .74 k= 1.00 k= .86
Demands carer's attention k = .86 k = .90 k= 1.00 k= .76
Non-algorithm items
Wanders off in public k = .79 k = .89 k = .89 k = .82
Destructive k = .57 k = .82 k = 1.00 k = .95
Noisy k = .85 k = .89 k = .97 k = .73
Has temper tantrums k = .87 k = .97 k = 1.00 k = .75
Aggressive k = .84 k = .93 k = 1.00 k = .85
Has objectionable personal habits k = .78 k = .78 k = 1.00 k = .88
Scatters or throws objects around k = 1.00 k = 1.00 k = 1.00 k = 1.00
Poor co-operation k = .89 k = .93 k = 1.00 k = .76
Needs constant supervision k = .79 k = .78 k = 1.00 k = .90
Demands carer's attention k = .86 k = .90 k = 1.00 k = .76
SLEEP PATTERN
Non-algorithm items
Has dif®culty falling asleep k = .90 k = .78 k = 1.00 k = .94
Has dif®culty remaining asleep k = .82 k = .81 k = 1.00 k = .84
Has night terrors or nightmares k = .93 k = .95 k = 1.00 k = .95
CATATONIC FEATURES
Algorithm items
Walks with head bowed and arms still k = .90 k = .90 k = 1.00 k = 1.00
Non-algorithm item
Hyperextends hands in unusual position k = .87 k = .94 k = .89 k = .90
Repeated approach and withdrawal 2 childr. 3 childr. 1 child 2 childr.
QUALITY OF SOCIAL INTERACTION
Algorithm item
Quality of social interaction r = .96 r = .96 r = .70 r = .88

Disco

Uploaded by

Copyright:

Available Formats

Disco

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Disco

Uploaded by

Copyright:

Available Formats

Journal of Child Psychology and Psychiatry 43:3 (2002), pp 307±325

The Diagnostic Interview for Social and

only. The Childhood Autism Rating Scale (CARS)

should proceed. Informants are encouraged to des-

Table 1 Characteristics of the children

Autistic low Learning Autistic high Language Typically

A Characteristics of school-aged children

Table 2 Number of DISCO ratings and level of inter-rater reliability

School-age children Pre-school children

Kappa or percentage Infancy, age of Untypical Infancy, age of Untypical

* Items in these sections rated twice ± `current' and `ever'.

Leekam, S.R. (1996). Features of autism in normally

book of autism and pervasive developmental disorders: Appendix

Manuscript accepted 22 October 2001

Inter-rater agreement for DISCO items

Appendix Table Continued

Setback in range of foods eaten 2 children 1 child

Appendix Table Continued

Lack of awareness of need for cleanliness k = .86 k = .86 NA NA

Appendix Table Continued

Appendix Table Continued

Appendix Table Continued

Special skills: Memory for routes 3 childr. 3 childr. k = .72 k = .72

Appendix Table Continued

You might also like