Journal of Pediatric Psychology, Vol. 12, No.

4, 1987

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
Cross-Disciplinary Perspectives on Autism1
Wendy L. Stone2
University of Miami, Mailman Center for Child Development, Miami, Florida

Received August 25, 1986; accepted October 31, 1986

A 23-item survey assessing knowledge and beliefs about autism was com-
pleted by 239 professionals in four disciplines: clinical psychology, pediatrics,
school psychology, and speech/language pathology. Their responses were
compared with those obtained from 18 specialists in the area of autism.
Whereas the specialists' views were consistent with those prevalent in the
research literature, individual disciplines displayed a number of misconcep-
tions regarding social, emotional, and cognitive aspects of the disorder. Some
misconceptions were specific to certain disciplines, while others were shared
by all groups. For example, only speech/language pathologists viewed au-
tism as an emotional disorder, whereas all four disciplines attributed un-
realistically high cognitive potential to autistic individuals. Diagnostic criteria
were also found to differ between the groups.
KEY WORDS: autism; survey; interdisciplinary.

One phenomenon associated with pediatric settings is the diversity of theo-

retical perspectives that can be found among different disciplines. Disparate
training backgrounds and professional orientations can lead to conflicting
views regarding assessment, diagnosis, and case management (Cerreto, 1980;
Drotar, 1983; Roberts & Wright, 1982). This phenomenon may become par-
ticularly problematic on large interdisciplinary teams, such as those involved

'The preparation of this paper was supported in part by the Florida Diagnostic and Learning
Resources System through a state general revenue appropriation for evaluation services in
exceptional student education.
2
A1J correspondence should be sent to Wendy L. Stone, University of Miami, Mailman
Center for Child Development, P.O. Box 016820, Miami, Florida 33101.
615

0I46-869VS7/12OO-O615SO4JJVO © 19*7 Plenum PuHntaj Q x p t n t k n

616 Stone

with developmental^ handicapped populations. Because of the complex na-

ture of handicapping conditions such as mental retardation, autism, and
cerebral palsy, pediatric psychologists often collaborate with a number of
different professionals on a single case (Mesibov, 1984).
Although the general issue of diverse training backgrounds and per-
spectives has been alluded to in the literature, very little has been written
about the effect of different training experiences on views and beliefs about

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
particular conditions. Consequently, the aim of this study was to determine
empirically the effects of different professional training backgrounds on
knowledge and attitudes regarding a specific developmental disorder. Au-
tism was chosen as the prototype for two principal reasons. First, it has been
observed that autistic children and their families commonly encounter a num-
ber of different professionals on their way to diagnosis and treatment (Farber
& Capute, 1984; Schopler, Mesibov, Shigley, & Bashford, 1984). This sug-
gests the relevance and potential importance of assessing views of autism
across a number of different disciplines. Second, autism is a disorder sur-
rounded by a great deal of diagnostic confusion. In the 40 years since its
identification, the increasing availability of empirical data has led to recon-
ceptualizations regarding fundamental aspects of the disorder, such as its
definition, etiology, and diagnosis (Schopler, 1983). Consequently, diverse
views of the disorder are likely to be found.
Cross-disciplinary views of autism were explored by having professionals
in four different disciplines complete a survey on autism. The disciplines of
clinical psychology, pediatrics, school psychology, and speech/language
pathology were chosen for their involvement with autistic children in diag-
nostic and/or treatment capacities. The survey was developed for use in this
study and designed to assess views regarding etiology, diagnosis, and specif-
ic features of the disorder. Documentation of differences between profes-
sional groups serves to highlight the importance of recognizing the individual
perspectives that each discipline brings to multidisciplinary pediatric settings.

METHOD
Survey Development

In order to examine views and beliefs about autism, a 23-item survey

was developed. The survey comprised two parts. Part I consisted of 21 state-
ments related to common misconceptions about autism, gleaned from the
literature (e.g., Farber & Capute, 1984; Schopler, 1983) as well as clinical
experience. The items were selected to represent three broad areas: social and
emotional features (8 items), cognitive characteristics (6 items), and general
descriptive features, including course and prognosis (7 items). Respondents
Perspectives on Autism (17

were asked to indicate the degree to which they agreed with each statement.
The ratings were obtained on a 6-point scale, with the following anchor points:
1 = fully agree, 2 = mostly agree, 3 = somewhat agree, 4 = somewhat
disagree, 5 = mostly disagree, and 6 = fully disagree.
Part II of the survey consisted of two questions focusing on diagnostic
criteria. For both questions, the same list of 18 behaviors or characteristics
was presented. The first question asked respondents to check all behaviors

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
that are required for a diagnosis of autism, and the second question asked
respondents to check all behaviors that are helpful, though not necessary,
in making the diagnosis. The list of behaviors included those required by
the Third Edition of the Diagnostic and Statistical Manual (DSM-HI; Ameri-
can Psychiatric Association, 1980) for diagnosis (e.g., lack of social respon-
siveness) as well as other characteristics found more commonly in nonautistic
populations (e.g., thought disorder).

Procedure

The survey was distributed to professionals in four disciplines: pedi-

atricians, clinical psychologists, speech/language pathologists, and school
psychologists. The disciplines were chosen to represent both community- and
school-based professionals whose typical involvement with autistic individuals
occurs at different points along the diagnosis-treatment continuum. The pedi-
atrician and clinical psychologist groups were drawn primarily from the com-
munity. Although both groups may be involved in the initial identification
of autism, only the clinical psychologists are likely to be involved in treat-
ment. The samples of school psychologists and speech pathologists were ob-
tained mostly from school settings. Whereas school psychologists are involved
primarily with diagnosis, speech pathologists are usually concerned only with
treatment.
All surveys were completed anonymously. Approximately half (49%)
were obtained via direct mailings, and the others (51%) were collected from
distribution at meetings or conferences. A total of 239 surveys were received:
48 from pediatricians, 97 from speech/language pathologists, 42 from clini-
cal psychologists, and 52 from school psychologists. The overall return rate
from direct mailings was 46% (31% for pediatricians, 45% for school psy-
chologists, 54% for speech/language pathologists, and 56% for clinical psy-
chologists). In order to assure that the clinical psychologists and speech
pathologists surveyed by mail were those who worked with children, tele-
phone queries were conducted prior to the mailings.
Since many of the survey items reflect opinions and attitudes rather
than absolutes, it was necessary to obtain a standard against which each dis-
cipline could be compared. Toward this end, the survey was also distributed
618 Stone

to a group of specialists in the field of autism. Specialists were defined as

individuals who had been directly and extensively involved in research or clin-
ical work in autism for a significant length of time (i.e., at least 5 years).
The specialists were culled from three university settings engaged in active
research with autistic populations: University of North Carolina, Yale Univer-
sity, and University of California at Los Angeles. Specialists from different
university centers were sought in order to minimize the potential biasing ef-

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
fect of setting-specific responses. The distribution procedure for this group
was somewhat different from that described above. In order to obtain
responses from specialists in the three settings, contact persons were asked
to direct the surveys to specific individuals meeting the "specialist" criteria.
Surveys were also completed anonymously by the specialists. Eighteen
surveys were collected (10 from UNC, 6 from UCLA, and 2 from Yale). The
specialist group was composed of the following disciplines: psychology
(44%), education (39%), and psychiatry (17%).

RESULTS
Survey Respondents

Demographic characteristics of the survey respondents are presented

in Table I. Professionals in all groups reported substantial experience in their

Table I. Work Characteristics of Survey Respondents

Respondents (No.)'
SPEC PED CLIN SCH LANG
Work variables (18) (48) (42) (52) (97)
Mean years in present
occupation 11.3 17.3 10.6 10.5 10.9
Work setting (%)
Private practice 5.6 77.1 59.5 11.5 6.2
University medical center 94.4 8.4 21.4 — 1.0
Community clinic _ 8.4 21.4 — —
Hospital _ 10.4 11.9 1.9 5.2
School 11.1 _ 11.9 96.2 85.5
College - - 9.5 - -
Work activities (%)
Direct care 72.2 97.9 88.1 96.2 95.9
Training 83.4 27.1 64.3 9.6 13.4
Research 83.3 8.3 40.5 1.9 5.1
Interdisciplinary
involvement (%) 50.0 18.8 47.6 44.1 18.5
"SPEC = autism specialists, PED = pediatricians, CLIN «• clinical psychologists,
SCH = school psychologists, LANG = speech/language pathologists.
Perspectives on Autism 619

fields 0-e., an average of at least 10 years). A one-way ANOVA yielded sig-

nificant group differences on this variable, F{4, 244) = 5.93, p < .0001;
post hoc comparisons revealed pediatricians to have worked significantly
longer in their present occupations than all other groups, /(240) = 2.54, p
< .05 for specialists; r(240) = 3.71, p < .001 for clinical psychologists; f(240)
= 4.02, p < .001 for school psychologists; /(240) = 4.31, p < .001 for speech

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
pathologists. The majority of respondents in all groups reported direct in-
volvement with patients or students, though different patterns of work ac-
tivities characterized individual disciplines.
Three demographic questions were designed to assess the respondents'
experience and familiarity with autism. Responses to these items are presented
in Table II. As expected, the specialists reported substantially greater ex-
perience with autistic individuals, relative to the other groups of profession-
als. A one-way ANOVA yielded significant group differences in the total
number of autistic individuals seen, F\4, 241) = 56.21, p < .0001; planned
comparisons revealed the group of specialists to have seen a significantly larger
number of autistic individuals than all other groups, /(237) = 13.54 for pedi-
atricians, 12.43 for clinical psychologists, 13.55 for school psychologists, and
14.36 for speech pathologists; all p < .001. No other group comparisons
were significant on this item.
Similar results were obtained on the question assessing the average num-
ber of autistic individuals seen per year. In all four professional groups, the
majority of respondents reported seeing three or fewer autistic individuals
in a typical year. In constrast, all of the specialists reported seeing at least
four per year, with the majority seeing over 10 autistic individuals in a typi-
cal year. Comparison of the respondents' degree of comfort using the diag-

Table D. Survey Respondents' Experience with Autism

Respondents (No.)'
SPEC PED CLIN SCH LANG
Experience variables (18) (48) (42) (52) (97)
Mean no. of all autistic
individuals seen 279.4 7.9 27.1 9.8 9.9
No. seen per year (%)
None 0 50.0 40.0 32.6 60.0
1-3 0 35.4 35.0 45.7 28.4
4-10 18.7 8.3 20.0 19.6 10.5
Over 10 81.3 6.3 •5.0 2.1 1.1
Mean degree of comfort
diagnosing individuals
as autistic 1.5 4.0 3.1 3.7 4.3
'SPEC = autism Specialists, ped = pediatricians, CLIN = clinical psychologists,
SCH = school psychologists, LANG = speech/language pathologists.
*LOWCT numbers reflect stronger agreement with the statement, "I feel comfortable
diagnosing or identifying a child as autistic."
620 Stone

nostic label of autism also revealed significant group differences, F\A, 254)
= 15.55, p < .0001). This item required respondents to rate their agree-
ment with the statement "I feel comfortable diagnosing or identifying a child
as autistic" on a scale from 1 to 6 (where 1 = fully agree and 6 = fully dis-
agree). It can be seen in Table II that the autism specialists were the only
group to agree strongly with this statement, reflecting comfort using the di-
agnosis. Planned comparisons revealed the specialists to feel significantly

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
more comfortable with the diagnosis than all other groups, f(250) = 5.97
for pediatricians, 3.79 for clinical psychologists, 5.28 for school psycholo-
gists, and 7.28 for speech pathologists; all p < .001.

Survey Results: Part I

Part I consisted of 21 statements, each of which was rated according

to degree of agreement (where 1 = fully agree, 2 = mostly agree, 3 = some-
what agree, 4 = somewhat disagree, 5 = mostly disagree, and 6 = fully
disagree). This response format was designed to force a choice between the
agree and disagree sides of the continuum. Mean scores of less than 3.5 can
be considered to reflect agreement with the statements, whereas scores above
3.5 indicate disagreement.
To evaluate the presence of group differences, separate one-way ANO-
VAs were calculated for each statement. Since the information of greatest
interest was the performance of each discipline relative to the specialists,
planned comparisons between each of the four_professional groups and the
specialist group were obtained for each item that yielded a significant ANO-
VA. (Two-tailed t tests were used for these analyses.) Due to the large num-
ber of comparisons involved, an effort was made to minimize the likelihood
of Type I errors by adopting a significance level of .01 for the overall F statis-
tic. Consequently, only those survey items yielding group differences at a
probability level of less than .01 on the ANOVA were subjected to the planned
comparisons.

Social and Emotional Features

Significant group differences were found for six of the eight items deal-
ing with social and emotional aspects of autism. The means for these items
and the results of the planned comparisons are presented in Table III.
For most items, planned comparisons revealed significant differences
between all disciplines and the specialists. The specialists strongly disagreed
with statements attributing emotional features to autism. In contrast, all four
Perspectives on Autism 621

Table m . Mean Ratings for Social-Emotional Items: Comparison of Each Discipline

to Specialists
Respondents*
Item SPEC PED CLIN SCH LANG
Autism is an emotional
disorder 5.71 4.25' 4.07* 3.76' 3.21'
Emotional factors play a

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
major role in the etiology of
autism 5.71 3.86' 4.52C 4.09' 3.48'
Autistic children are
deliberately negativistic and
noncompliant 5.22 4.25' 5.00 4.80 4.48*
It is difficult to distinguish
between autism and childhood
schizophrenia 4.78 2.81' 3.98* 3.25' 3.28'
Autistic children do not show
social attachments, even to
parents 5.22 2.88' 3.6C 3.05' 3.24'
Autistic children do not show
affectionate behavior 4.89 3.35' 3.85' 3.08' 3.61'
"SPEC = autism specialists, PED = pediatricians, CLIN = clinical psychologists,
SCH = school psychologists, LANG = speech/language pathologist!.
V < 05.
'D < .01.
*p < .001.

disciplines were more likely to think of autism as an emotional disorder, t(246)

= 3.55 for pediatricians, 3.91 for clinical psychologists, 4.76 for school psy-
chologists, and 6.53 for speech pathologists; and to see emotional factors
as causing autism, /(248) = 4.91 for pediatricians, 3.09 for clinical psychol-
ogists, 4.35 for school psychologists, and 6.37 for speech pathologists. All
four disciplines also found it more difficult to differentiate between autism
and childhood schizophrenia, compared with the specialists, /(250) = 5.25
for pediatricians, 2.10 for clinical psychologists, 4.13 for school psycholo-
gists, and 4.30 for special pathologists. Only the speech pathologists (t(25l)
= 2.39) and the pediatricians (r(251) = 2.93) felt more strongly than the
specialists that autistic children are intentionally negativistic.
The two social items also revealed significant differences between the
specialists and all other groups. Relative to the specialists, all disciplines were
more likely to believe that autistic children do not show any social attach-
ments, /(251) = 6.24 for pediatricians, 4.25 for clinical psychologists, 5.83
for school psychologists, and 5.69 for speech pathologists; or affectionate
behavior, t{225) = 3.62 for pediatricians, 2.71 for clinical psychologists, 4.88
for school psychologists, and 3.68 for speech pathologists.
622 Stone

Cognitive Features

Five of the six statements focusing on cognitive aspects of autism rev-

ealed significant group differences. Table IV presents the means for each
item, as well as the results of the planned comparisons.
The two questions related to intelligence level revealed significant differ-
ences between all disciplines and the specialists. Whereas the specialists agreed

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
that most autistic children are mentally retarded, and disagreed that they are
more intelligent than approapriate testing indicates, opposite results were ob-
tained from the four disciplines (for the former item, t(249) = 6.22 for pedi-
articians, 5.78 for clinical psychologists, 6.26 for school psychologists, and
8.86 for speech pathologists; for the latter item, f(242) = 7.47 for pediatri-
cians, 5.85 for clinical psychologists, 7.01 for school psychologists, and 8.09
for speech pathologists). All disciplines also viewed autistic children as more
untestable, t(249) = 5.96 for pediatricians, 3.15 for clinical psychologists,
3.24 for school psychologists, 4.93 for speech pathologists, compared with
the specialists.
With respect to specific cognitive abilities, all disciplines except clini-
cal psychologists were more likely to view autistic children as possessing spe-
cial talents, t(249) = 3.95 for pediatricians, 3.16 for school psychologists,
and 4.75 for speech pathologists than the specialists. Only the pediatricians
differed from the specialists, f(251) = 3.08, by agreeing with the assertion
that most autistic children do not speak.

Table IV. Mean Ratings for Cognitive Items: Comparison of Each Discipline to
Specialists
Respondents
Item SPEC PED CLIN SCH LANG
Most autistic children do
not talk 4.00 2.83* 3.95 3.72 3.57
Most autistic children have
special talents or abilities 4.78 3.38* 4.26 3.67' 3.21'
Most autistic children are
also mentally retarded 1.17 3.66C 3.52' 3.65' 4.46'
Autistic children are more
intelligent than scores from
appropriate tests indicate 4.89 2.48 s 2.98* 2.65' 2.48*
Autistic children are
untestable 5.83 3.98* 4.83' 4.83* 4.41'
"SPEC - autism specialists, PED =• pediatricians, CLIN = clinical psychologists,
SCH = school psychologists, LANG = speech/language pathologists.
b
p < .01.
c
p < .001.
Perspectives on Autism 623

Descriptive Features.

Items in this category included general descriptive characteristics of au-

tism as well as those related specifically to the course and prognosis of the
disorder. Five of the seven items revealed significant group differences (see
Table V).

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
All disciplines viewed autism as a somewhat more temporary condi-
tion than did the specialists; they were more likely to believe that autism ex-
ists only in childhood, t(249) = 3.95 for pediatricians, 2.94 for clinical
psychologists, 3.79 for school psychologists, and 2.76 for speech patholo-
gists; and that autism is a condition that can be outgrown, /(247) = 3.97
for pediatricians, 2.91 for clinical psychologists, 2.82 for school psycholo-
gists, 4.66 for speech pathologists. All disciplines also felt less strongly than
the specialists that autism is a developmental disorder, t(240) = 5.48 for pedi-
atricians, 2.58 for clinical psychologists, 3.83 for school psychologists, and
6.82 for speech pathologists.
The group of specialists strongly disagreed with the assertion that au-
tism is more common among upper SES and educational levels. Every dis-
cipline's response to this item reflected less disagreement than the specialists;
/(244) = 6.01 for pediatricians, 5.79 for clinical psychologists, 4.12 for school
psychologists, and 4.59 for speech pathologists. The only other group differ-
ences in this area occurred with respect to prognostic issues. Clinical psy-

Table V. Mean Rating for Descriptive Items: Comparison of Each Discipline to

Specialists
Respondents"
Item SPEC PED CLIN SCH LANG
Autism exists only in
childhood 5.94 4.35' 4.74' 4.43' 4.91'
With the proper treatment.
most autistic children
eventually outgrow autism 5.72 4.52' 4.83' 4.88' 4.42'
Autism is a developmental
disorder 1.00 3.3C 2.10* 2.59' 3.64'
Autism occurs more commonly
among higher sodoeconomic
and educational levels 5.67 3.33' 3.37' 4.07' 4.01'
Even with early intervention,
the prognosis for independent
community functioning of
autistic individuals is poor 3.33 2.77 2.43' 2.72 3.15
"SPEC = autism specialists, PED = pediatricians, CLIN = clinical psychologists,
SCH = school psychologists, LANG = speech/language pathologists.
*/> < .05.
c
p < .01.
*p < .001.
624 Stone

chologists were significantly more pessimistic than the specialists regarding

prognosis for independent community functioning, 7(249) = 2.76.

Survey Results: Part II

Part II questions related most directly to diagnostic criteria. The first

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
item asked respondents to check which of 18 characteristics or behaviors are
required for a diagnosis of autism, and the second asked respondents to check
which characteristics are helpful, though not necessary, in diagnosing au-
tism. Since the specialists' responses were used as a standard for compari-
son, it is important to understand the diagnostic criteria employed by this
group. All characteristics endorsed as either required or helpful by more than
half of the specialists are presented in Table VI.
In Table VI the specialists' responses can also be compared with the
diagnostic criteria provided in DSM-III (American Psychiatric Association,
1980). Primary characteristics are those described in DSM-III as necessary
for the diagnosis of autism; secondary characteristics are those that are men-
tioned as examples but are not required for diagnosis. As Table VI reveals,
there was substantial agreement between the specialists and DSM-III; all of
the primary characteristics listed in DSM-III were endorsed as required by
the majority of the specialists, and all of the secondary characteristics were
endorsed as helpful by the majority.
In order to examine group differences in diagnostic criteria used, a one-
way ANOVA was performed for each item; again a significance level of .01
was adopted. Results of the ANOVA and subsequent planned comparisons
are presented in Table VII.

Table VI. Charactcrsitks Endorsed by Majority of Specialists: Comparison with DSM-III

Criteria Specialists (°h)
Specialists (%)
Characteristics Required Helpful DSM-III*
Language delays 89 50 P
Onset of symptoms before 30 months 89 56 P
Lack of social responsiveness 83 56 P
Need for sameness; resistance to change
in routine 61 67 P
Peculiar speech characteristics 56 78 P
Lack of eye contact 11 83 S
Unusual mannerisms such as finger
flicking 11 83 S
Rigid or stereotyped play activities 22 78 —
Unusual sensory responses 33 78 S
Mutism 0 67
s
"P = primary characteristics, S «• secondary characteristics.
Perspectives on Autism 625

Table VTI. Mean Endorsement of Characteristics as Required or Helpful: Comparison

of Each Discipline to Specialists
Respondents (
Characteristics SPEC PED CLIN SCH LANG
Required
Onset before 30 months 89 60* 81 72 40"
Unusual sensory responses 33 29 41 58 56

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
Rigid or stereotyped play 22 81" 48 64' 60"
Lack of eye contact 11 79" 31 64" 74'
Attention deficits 11 29 38' 58" 69"
Unusual mannerisms 11 48 s 41' sr 66"
Inappropriate laughing or
giggling 6 21 12 25 45"
Mutism 0 56" 17 17 30*
Sudden, unexpected mood
changes 0 35' 14 29* 49"
Aggressive behavior 0 8 2 14 21'
Helpful
Hallucinations 0 6 24' 4 12
Thought disorder 0 25' 31' 25' 40"
"SPEC = Autism specialists, PED = Pediatricians, CLIN = Clinical psychologists,
SCH = School psychologists, LANG •= Speech/Language pathologists.
*p < .05.
c
p < .01.
i
p < .001.

With respect to required characteristics, 10 items reflected significant

group differences. Planned comparisons revealed pediatricians to be less likely
than specialists to endorse early onset of symptoms, t(252) = 2.23; and more
likely to require rigid/stereotyped play activities, /(252) = 4.52; lack of eye
contact, t{252) = 5.57; unusual mannerisms such as finger flicking, f(252)
= 2.76; mutism, r(252) = 4.79; and sudden, unexplained mood changes,
t{252) = 2.85, relative to the specialists. Clinical psychologists differed from
the specialists only in their higher endorsement of attention deficits, t(252)
= 2.05, and unusual mannerisms, t(252) = 2.16. School psychologists were
more likely than the specialists to require stereotyped play activities, t{252)
= 3.19; lack of eye contact, /(252) = 4.33; attention deficits, t(252) = 3.65;
unusual mannerisms, /(252) = 3.09; and sudden mood changes, t{252) =
2.34, for a diagnosis of autism. Speech pathologists differed from the
specialists on 9 of the 10 items. They were less likely to use early age of onset
as a criterion, f(252) = 4.11; and more likely to require stereotyped play,
f(252) = 3.10; lack of eye contact, f(252) = 5.56; attention deficits, t{252)
= 4.83; unusual mannerisms, r(252) = 4.43; inappropriate laughing or gig-
gling, t(252) = 3.59; mutism, r(252) = 2.74; sudden mood changes, t{252)
= 4.20; and aggressive behavior, f(252) = 2.47. Early age of onset was the
only DSM-III primary characteristic that revealed group differences.
626 Stone

The second question yielded significant group differences on only two

additional characteristics. Clinical psychologists were the only group to en-
dorse hallucinations as significantly more helpful in diagnosing autism com-
pared with the specialists, /(252) = 2.81. All groups were more likely to
consider thought disorder to be helpful in diagnosing autism relative to the
specialists, ^(252) = 2.01 for pediatricians, 2.44 for clinical psychologists,
2.03 for school psychologists, and 3.48 for speech pathologists. It should

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
be noted, however, that the rates of endorsement for most groups were rela-
tively low on this item.

DISCUSSION

Results on the first part of the survey reveal misconceptions about var-
ious aspects of autism to exist in all professional groups. The specialists'
responses reflect the current trends and conceptualizations prevalent in the
literature. In many cases, the differences found between individual disciplines
and the specialists can be attributed to the persistence of older views of au-
tism that have since been challenged or refuted.
Early conceptualizations placed a heavy emphasis on the role of emo-
tional factors in the development and course of autism. Autism was described
as the earliest form of childhood schizophrenia (Kanner, 1949), and believed
to be caused by psychogenic factors such as parental pathology (e.g., Bettel-
heim, 1967). However, these views have been largely abandoned, in favor
of evidence supporting underlying organic (e.g., genetic, neurological) fac-
tors (see DeMyer, Hingtgen, & Jackson, 1981, and Morgan, 1986, for more
complete discussions of this issue). This shift in thinking is evidenced in the
changes made in the third revision of the Diagnostic and Statistical Manual
(DSM-III), where autism is listed as a developmental disorder and explicitly
differentiated from schizophrenia. In previous editions, "childhood
schizophrenia" was the only category available for such children.
Consistent with current views, the autism specialists in this study re-
jected notions of emotional involvement, while endorsing the view of au-
tism as a developmental disorder. Most of the other disciplines share these
beliefs. However, the speech/language pathologists continue to see autism
as an emotional disorder. Their responses on Part II also reflect this per-
spective; almost half of this group consider sudden, unexplained mood
changes and inappropriate laughing and giggling to be required behaviors
for a diagnosis of autism.
Major shifts in thinking have also occurred with respect to intellectual
functioning and other cognitive aspects of autism. Kanner (1943) originally
believed autistic children to have normal intellectual potential. This view seems
Perspectives on Autism 627

to have been derived, at least in part, from Kanner's observation of peak

cognitive skills in his sample. However, empirical research over the past 40
years has not supported the popular notions that most autistic children have
average intellectual abilities or possess special talents. On the contrary, studies
have repeatedly found that the majority (about 70%) of autistic individuals
function within the mentally retarded range (Dawson & Mesibov, 1983). Fur-
thermore, contrary to the idea of autistic children being "untestable," IQ

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
scores have been demonstrated to provide reliable and valid measurement
of intellectual abilities in this population (Marcus & Baker, 1986; Rutter,
1983).
Again, the specialists' responses were consistent with these views.
However, every discipline was found to harbor some misconceptions in this
area. All four groups demonstrated a belief in average intellectual function-
ing and the inability of intelligence tests to tap the true potential of the au-
tistic individual. Pediatricians and speech/language pathologists go one step
further, in regarding autistic children as possessing special talents or abili-
ties. The presumption of normal intelligence may have the most serious im-
plications for those disciplines involved in assessment or teaching activities.
Failure to consider the possibility of mental retardation may lead to overly
high expectations and repeated failure experiences.
Another aspect of cognitive functioning is that of language. One of the
numerous language problems often associated with autism is the failure to
develop meaningful speech. This occurs in approximately half of all autistic
children (Rutter, 1978). The group of pediatricians viewed lack of speech
as a more frequent concomitant of autism than it appears to be. They en-
dorsed the assertion that most autistic children do not talk. Furthermore,
results on Part II reveal that the majority of pediatricians require mutism
for a diagnosis of autism. Assuming continuity between beliefs and prac-
tice, this suggests that half of the autistic population (i.e., those with speech)
may go unrecognized by many pediatricians.
Another early view that has been challenged is the belief that autistic
children are found more commonly in upper-class families. Kanner's (1943)
original observation to this effect was followed by other early papers report-
ing the same phenomenon (see Steinhausen, Gobel, Breinlinger, & Wohlle-
ben, 1986, and Wing, 1980 for reviews). However, a number of selection
factors are now felt to have biased these findings, and more recent epidemi-
ological studies have found autism to occure at all socioeconomic levels (Scho-
pler, 1983). The pediatricians and clinical psychologists are the only groups
who continue to endorse the outdated notion of overrepresentation in higher
socioeconomic classes.
The nature of the social impairments in autism has been the subject
of more recent controversy. While there seems to be general agreement that
«8 Stone

social deficits represent a central feature of autism, the precise nature of these
social problems has been the focus of few empirical investigations (Howlin,
1986; Schopler & Mesibov, 1986). Thus far, a great deal of variability in the
expression of social deficits in autistic individuals has been demonstrated.
Social responsiveness can vary from one autistic individual to the next, and
social interest often increases with age (Fein, Pennington, Markowitz, Braver-
man, & Waterhouse, 1986; Howlin, 1986). Furthermore, recent reports have

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
described a broader range of social abilities in autistic individuals than has
previously been assumed. Autistic individuals have been found to demon-
strate attachment and affectionate behaviors and to form social relation-
ships (Cohen, Volkmar, & Paul, 1986; Sigman & Ungerer, 1984; Volkmar,
Cohen, & Paul, 1986).
The specialists' responses reflect this newer perspective. They recognize
that autistic children may show social attachments and affectionate behaviors.
Responses of the clinical psychologist group were in the same direction, but
less extreme. However, the pediatricians, school psychologists, and
speech/language pathologists see autistic children as significantly more so-
cially restricted and less variable in their social behaviors. Although the ac-
tual consequences of such a perspective cannot be predicted, it seems possible
that the sight of a child clinging to his/her parent may lead some profession-
als to reject autism as a potential diagnosis.
The second part of the survey yields results more directly related to di-
agnosis. Professionals in different disciplines employ different criteria in
diagnosing autism. The specialists' responses were quite consistent with the
specifications in DSM-III. Few group differences were found on those charac-
teristics considered primary by DSM-III. The only exception was that pedi-
atricians and speech pathologists were less likely to cite early age of onset
as required for diagnosis.
On the other hand, in many cases individual disciplines attribute great-
er importance to certain characteristics relative to the specialists. The charac-
teristics of rigid play activities and lack of eye contact provide the most
dramatic examples; both were endorsed as helpful characteristics by the
majority of specialists, but only a very small proportion of specialists con-
sidered them necessary for a diagnosis. In contrast, the majority of pediatri-
cians, school psychologists, and speech/language pathologists rated these
characteristics as required. The research literature supports the specialists'
views. Although stereotyped play activities (Volkmar et al., 1986) and lack
of eye contact (Dawson & Galpert, 1986; Howlin, 1986) are commonly ob-
served behaviors in autism, they are associated with considerable intra- and
interindividual variability, and consequently cannot be diagnostic requisites.
Other discrepancies of this type include pediatricians' overreliance on mutism
Perspectives on Autism 629

as a required characteristic and the endorsement of attention deficits and

unusual mannerisms as required by school psychologists and speech/language
pathologists. It can easily be seen how the use of such disparate criteria for
diagnosing autism might impede the attainment of diagnostic consensus on
an interdisciplinary team.
Although the results obtained in this study clearly suggest that inter-
disciplinary differences in views of autism do exist, it must be remembered

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
that direct comparisons between the four disciplines were not made. Rather,
it was deemed more meaningful to have a frame of reference against which
each group's responses could be compared. Although the reference group
of specialists was relatively small, it was composed of individuals drawn from
several settings, with the idea of representing a broad range of views. The
adequacy of this comparison group is demonstrated by the high degree of
consistency between the specialists' responses and the views currently dominat-
ing the literature.
Unfortunately, it was impossible to match the groups on demographic
variables. Consequently, the extent to which group differences reflect dis-
ciplinary differences independent of experiential factors cannot be deter-
mined. In particular, it is difficult to tease out effects due to training
background and work variables, since the two covary naturally (e.g., school
psychologists tend to work in school settings, and speech pathologists tend
to be involved in treatment activities, etc.). For example, the similarities
between the responses of the clinical psychologists and the specialists may
be explainable in terms of demographic variables. Relative to the other dis-
ciplines, they were more similar to the specialists in work and experience vari-
ables: A larger portion of clinical psychologists worked in university settings
and were involved in research, and the clinical psychologist group had also
seen more autistic children. Of course, similarities may also have been stron-
gest because 44% of the specialists were clinical psychologists! Future research
might overcome this problem by surveying a larger number of professionals
who actually work in interdisciplinary settings and by obtaining a specialist
group composed of individuals from the same backgrounds as the survey
respondents.
Despite the above limitations, this study presents support for the con-
tention that perspectives on a singular disorder can differ from one discipline
to the next. Aside from providing information specific to autism, the impli-
cations of the present results extend to any professional working in an inter-
disciplinary setting. Recognition of cross-disciplinary differences may be the
first step toward transcending these differences. Continued research in this
area may provide a fruitful approach toward facilitating more productive
and compatible interdisciplinary liaisons.
630 Stone

REFERENCES

American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disord-
ers (3rd ed.). Washington, DC: Author.
Bettelheim, B. (1967). Theempty fortress—Infantile autism and the birth of the self. New York:
Free Press.
Cerreto, M. C , (1980). Training issues: The pediatric psychologist-child neurologist team. Journal
of Pediatric Psychology, 5, 253-261.

Downloaded from http://jpepsy.oxfordjournals.org/ at University of California, San Francisco on January 29, 2015
Cohen, D. J., Volkraar, F. R., & Paul, R. (1986). Issues in the classification of pervasive de-
velopmental disorders: History and current status of nosology. Journal of the American
Academy of Child Psychiatry, 25, 158-161.
Dawson, O., & Galpert, L. (1986). A developmental model for facilitating the social behavior
of autistic children. In E. Schopler & G. B. Mesibov (Eds.), Social behavior in autism
(pp. 237-261). New York: Plenum Press.
Dawson, G., & Mesibov, G. B. (1983). Childhood psychoses. In C. E. Walker & M. C. Roberts
(Eds.), Handbook of clinical child psychology (pp. 543-572). New York: Wiley.
DeMyer, M. K., Hingtgen, J. N., & Jackson, R. K. (1981). Infantile autism reviewed: A de-
cade of research. Schizophrenia Bulletin, 7, 388-451.
Drotar, D. (1983). Transacting with physicians: Fact and fiction. Journal of Pediatric Psychol-
ogy, 8, 117-127.
Farber, J. M., & Capute, A. J. (1984). Understanding autism. Clinical Pediatrics, 23, 199-202.
Fein, D., Pennington, B., Markowitz, P., Bravennan, M., & Waterhouse, L. (1986). Toward
a neuropsychological model of infantile autism: Are the social deficits primary? Journal
of the American Academy of Child Psychiatry, 25, 198-212.
Howlin, P. (1986). An overview of social behavior in autism. In E. Schopler & G. B. Mesibov
(Eds.), Social behavior in autism (pp. 103-131). New York: Plenum Press.
Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child, 2, 217-250.
Kanner, L. (1949). Problems of nosology and psychodiagnostks of early infantile autism. Ameri-
can Journal of Orthopsychiatry, 19, 416-426.
Marcus, L. M., & Baker, A. F. (1986). Assessment of autistic children. In R. J. Simeonsson
(Ed.), Psychological assessment of special children (pp. 279-304). Boston: Allyn & Bacon.
Mesibov, G. B. (1984). Evolution of pediatric psychology: Historical roots to future trends.
Journal of Pediatric Psychology, 9, 3-11.
Morgan, S. B. (1986). Early childhood autism: Changing perspectives. Journal of Childhood
and Adolescent Psychotherapy, 3, 3-9.
Roberts, M. C , & Wright, L. (1982). The role of the pediatric psychologist as consultant to
pediatricians. In J. M. Tuma (Ed.), Handbook for the practice of pediatric psychology
(pp. 251-288). New York: Wiley.
Rutter, M. (1978). Diagnosis and definition. In M. Rutter <fc E. Schopler (Eds.), Autism: A
reappraisal of concepts and treatment (pp. 1-25). New York: Plenum Press.
Rutter, M. (1983). Cognitive deficits in the pathogenesis of autism. Journal of Child Psycholo-
gy and Psychiatry, 24, 513-531.
Schopler, E. (1983). New developments in the definition and diagnosis of autism. Advances
in Clinical Child Psychology, 6, 93-127.
Schopler, E., & Mesibov, G. B. (1986). Introduction to social behavior in autism. In E. Scho-
pler and G. B. Mesibov (Eds.), Social behavior in autism {pp. 1-11). New York: Plenum
Press.
Schopler, E., Mesibov, G. B., Shigley, R. H., & Bashford, A. (1984). Helping autistic children
through their parents: The TEACCH model. In E. Schopler & G. B. Mesibov (Eds.),
The effect of autism on the family (pp. 65-81). New York: Plenum Press.
Sigman, M., & Ungerer, J. A. (1984). Attachment behaviors in autistic children. Journal of
Autism and Developmental Disorders, 14, 231-244.
Steinhausen, H., Gobel, D., Breinlinger, M., & Wohlleben, B. (1986). A community survey
of infantile autism. Journal of the American Academy of Child Psychiatry, 25, 186-189.
Volkmar, F. R., Cohen, D. J., & Paul, R. (1986). An evaluation of DSM-III criteria for infan-
tile autism. Journal of the American Academy of Child Psychiatry, 25, 190-197.
Wing, L. (1980). Childhood autism and social class: A question of selection? British Journal
of Psychiatry. 137. 410-417.