Content uploaded by Chris Oliver
Author content
All content in this area was uploaded by Chris Oliver
Content may be subject to copyright.
J Abnorm Child Psychol (2007) 35:17–28
DOI 10.1007/s10802-006-9077-0
ORIGINAL PAPER
Is Theory of Mind Understanding Impaired in Males with Fragile
X Syndrome?
Cathy M. Grant · Ian Apperly · Chris Oliver
Published online: 23 November 2006
C
Springer Science+Business Media, LLC 2006
Abstract Males with fragile X syndrome (FXS) have dif-
ficulties with social interaction and many show autistic fea-
tures. This study examined whether the social deficits charac-
teristic of FXS are associated with theory of mind difficulties.
Two groups of boys with FXS participated: a group with few
autistic features and a group with many autistic features. An
intellectual disability control group also participated. In ad-
dition to using standard theory of mind tasks, new techniques
were used that were able to separate out the various process-
ing demands of the task (e.g., memory, inhibitory control).
Overall, the findings indicate that both groups of boys with
FXS have difficulty with theory of mind tasks compared to an
intellectual disability control group. However, both groups
with FXS also performed worse on comparison trials that
required working memory but not theory of mind. Theory
of mind difficulties are likely to be an important aspect of
the FXS clinical profile, but are most likely the result from a
more basic difficulty with working memory.
Keywords Fragile X syndrome
.
Autistic spectrum
disorder
.
Theory of mind
.
Inhibition
.
Working memory
Introduction
Fragile X syndrome (FXS) is a well-recognized cause of in-
tellectual disability and developmental delay in males and
C. M. Grant (
)
Paediatric Psychology, Child Development Centre, Windsor
Building, Leicester Royal Infirmary,
Leicester, LE1 5WW, United Kingdom
C. M. Grant · I. Apperly · C. Oliver
School of Psychology, University of Birmingham, Edgbaston,
Birmingham, B15 2TT United Kingdom
females. Numerous studies have attempted to identify a spe-
cific cognitive-behavioural phenotype associated with FXS
(reviewed in Bennetto & Pennington, 2002). In affected
males over 90% will present with intellectual disability usu-
ally in the mild-moderate range, compared with the majority
of females with full mutation FXS who have IQs that fall
within the low average range of normal ability (Bennetto
& Pennington, 2002). Although FXS is present at birth, its
behavioural manifestations are often not apparent until late
in the first year of life. The fragile-X behavioural pheno-
type is still being defined but notable characteristics include
social difficulties (e.g. deficits in social interactions with
peers and social withdrawal) and autistic-like behaviours
that include poor eye contact, stereotypic movements and
odd communication, including echolalia and perseverative
speech (e.g. Belser & Sudhalter, 2001; Sudhalter & Belser,
2001).
Much of the research into cognitive impairment in FXS
has focused upon executive function. Executive function
includes processes such as maintaining and updating in-
formation in working memory, cognitive flexibility, plan-
ning, initiation, inhibition and attentional regulation, in short,
abilities that are engaged in the generation and monitor-
ing of goal-directed behaviour. Numerous studies have re-
ported impaired executive functioning abilities in males
and females with FXS (e.g., Bennetto, Pennington, Porter,
Taylor, & Hagerman, 2001; Loesch, Bui, Grigsby, Butler,
Epstein, Huggins et al., 2003). Studies of children with
autism and typically developing children have shown ex-
ecutive functioning to be related to social cognitive abili-
ties and in particular, theory of mind abilities (e.g., Russell,
1997; Perner & Lang, 2000; Carlson & Moses, 2001). More
recently, researchers have also begun to examine whether in-
dividuals with FXS might also have impaired theory of mind
abilities.
Springer
18 J Abnorm Child Psychol (2007) 35:17–28
Theory of mind and Fragile X
It is increasingly recognised that reasoning about mental
states such as beliefs, desires and knowledge (often referred
to as “theory of mind”) is central to a range of social-
cognitive activities including the ability to communicate and
to explain and predict behaviour (Astington, 2000; Sperber,
2000). A task commonly used to examine mental state rea-
soning in children requires the inference that someone has
a false belief (Wimmer & Perner, 1983). For example, the
child might be told a story where Simon puts his chocolate in
the cupboard, then goes outside to play. While he is away his
mother moves the chocolate to the fridge. The child is asked
where Simon will first look for his chocolate when he returns.
To answer correctly, they must infer that Simon thinks that
the chocolate is still in the cupboard. Manyfour-year-olds an-
swer correctly, while many three-year-olds judge incorrectly
that he will look in the fridge (i.e., they answer from their
own knowledge and not the perspective of the other person).
Distinctive impairment on false belief, and other theory of
mind tasks, is often also observed in individuals with autism
and it has been suggested that this deficit could account for
the triad of impairments in socialisation, communication and
imagination (Baron-Cohen, 2000; Yirmiya, Erel, Shaked, &
Solomonica-Levi, 1998). It is estimated that 15–25% of boys
with FXS meet the diagnostic criteria for autism (e.g., Bailey
et al., 1998; Dykens & Volkmar, 1997). An important issue
for studies of theory of mind in FXS is whether any deficit
in theory of mind is an artifact of the high co-morbidity be-
tween FXS and autism or if such deficits are characteristic
of FXS irrespective of autistic features. Several studies have
found that individuals with intellectual disability but with-
out autism also perform poorly on false belief tasks (Yirmiya
et al., 1998). The issue of whether the poor performance of
individuals with intellectual disability reflects theory of mind
deficits or difficulties with task demands is unresolved.
Three research groups have investigated theory of mind
ability in people with FXS, yielding inconsistent results.
Mazzocco, Pennington, and Hagerman (1994) compared 19
adult females with FXS with a group of 27 women carry-
ing the fragile X gene and a group of 56 control women
who tested negative for the gene. All participants had IQ’s
of 70 or above and no participants were reported to have a
diagnosis of autism. The study found that performance on a
perspective-taking task was related to IQ rather than fragile
X group status. However, the perspective-taking task used
in this study was originally designed for children so may
have lacked the sensitivity to detect subtle deficits in the-
ory of mind in adult participants. Garner, Callias, and Turk
(1999) addressed theory of mind in a group of eight boys
with FXS (without autism) and reported that significantly
more children with FXS failed a first order false belief task
(Smarties task) than a comparison group of eight children
with intellectual disability of unknown aetiology matched
on chronological age and verbal mental age. However, on
another first order false belief task (the Sally-Anne task)
and a second-order false belief task there were no statisti-
cally significant group differences. Only seven of the eight
children in each group completed the Sally-Anne task, of
which five boys in the FXS group failed the task, compared
to one boy in the control group. The authors suggested that
their findings did not allow them to assess whether theory
of mind deficits were a specific feature of the syndrome
since performance on one of the tasks was related to overall
level of ability. A third study by Cornish, Burack, Rahman,
Munir, Russo, & Grant (2005) compared the performance
of children with FXS (without autism) with children with
Down’s syndrome matched on chronological age and verbal
mental age on two trials of the Sally-Anne false belief task.
The children with FXS performed similarly to children with
Down’s syndrome, with just under half the children in each
group passing the task. The authors concluded that while
children with FXS did have difficulties with theory of mind,
the deficit was not as severe as that previously reported for
children with autism, who would have been expected to per-
form less well than children with Down’s syndrome on false
belief tasks. Taken together, these studies suggest that chil-
dren with FXS often do make errors on false belief tasks.
However, it is unclear whether this pattern is a distinctive
feature of the FXS cognitive phenotype, or whether it only
reflects the intellectual disability that is commonly apparent
in children with FXS. A further issue is that the severity of
autistic features in the samples employed in Garner et al.
(1999) and Cornish et al. (2005) was not reported. Although
these studies included participants with FXS who did not
have a diagnosis of autism, some measure of autistic fea-
tures would have been useful since boys with FXS who do
not meet the diagnostic criteria for autism often still display
significant autistic features (Dykens & Volkmar, 1997).
Why might individuals with FXS fail false belief tasks?
The rationale for the present investigation is that, besides
the inconsistent findings in existing studies, there are also
methodological reasons why these studies might not be able
to give a clear picture of the existence or the origins of theory
of mind difficulties in FXS. First, the standard false belief
tasks used inthese, and most other studies of belief reasoning,
require participants both to reason about false beliefs (that
Simon thinks the chocolate is still in the cupboard) and to
resist interference from their own knowledge (that in fact the
chocolate is now in the fridge). The need to resist inference
places demands upon executive function. This confound is
particularly relevant because children with FXS have well-
attested impairments in executive function that are dispro-
portionate to their general level of intellectual impairment
Springer
J Abnorm Child Psychol (2007) 35:17–28 19
(Cornish, Sudhalter, & Turk, 2004). If children with FXS
have disproportionate difficulty resisting interference from
their own perspective, this alone could cause them to per-
form less well on standard false belief tasks than comparison
groups matched simply on general intellectual disability.
A second reason why it is difficult to reach clear con-
clusions from existing studies of false belief reasoning in
children with FXS is that false belief tasks make substantial
demands on working memory. Although the standard false
belief tasks used by Garner et al. (1999) and Cornish et al.
(2005) included check questions to ensure that the participant
could remember crucial facts about the story, remembering
these facts is not the only demand that the false belief task
makes on working memory. It is also necessary to remember
the sequence of events, which is not measured adequately by
check questions. Since children with FXS have impairments
in working memory that are disproportionate to their general
level of intellectual impairment (Munir, Cornish, & Wilding,
2001), it could be that any difficulty they have on false belief
tasks arises from working memory impairment, rather than
a more specific “theory of mind” impairment.
These considerations suggest two distinct questions about
whether theory of mind problems are part of the cognitive
phenotype of FXS. One question is whether children with
FXS are impaired on theory of mind tasks (such as false
belief tasks), in comparison to children with a similar gen-
eral level of intellectual ability. This is an important clinical
question addressed (albeit inconclusively) in existing stud-
ies. The second question concerns the nature of any theory
of mind deficit. Some authors propose that theory of mind
depends upon a dedicated cognitive system with a discrete
neural basis (for a discussion of this issue see e.g., Apperly,
Samson, & Humphreys, 2005; Frith & Frith, 2003; Saxe,
Carey, & Canwisher, 2004). Theory of mind impairment in
FXS might be the result of damage to such a system. It
is also widely recognised that theory of mind abilities de-
pend critically upon more general cognitive processes, such
as working memory and inhibitory control. Thus, theory of
mind impairment in FXS may be the indirect result of a
primary impairment to these processes.
The present study
The present study employed two first order standard false
belief tasks, plus two versions of a video-based, non-verbal
false belief reasoning task. Inclusion of two first order stan-
dard false belief tasks (the location-change task and the de-
ceptive box task) allowed us to gauge the performance of
children in this sample in comparison with a large number
of previous studies. The standard tasks included memory
check questions within each trial, though for the reasons al-
ready discussed, these were considered a rather weak test of
the contribution of working memory problems to the errors
children with FXS might make on false belief tasks. Two ver-
sions of the video-based tasks were used, “reality known”
and “reality unknown”. Inclusion of the video-based tasks
allowed us to address our first question by assessing be-
lief reasoning over a wider range of tasks than in previous
studies.
Importantly, the video-based tasks gave us two ways to
investigate whether individuals with FXS showed a primary
ToM deficit, or a deficit that was a consequence of impair-
ment to more general cognitive resources such as working
memory and inhibitory control. First, they included separate
comparison trials that did not require belief reasoning, but
that made demands on working memory that corresponded
to specific incidental demands of the false belief trials. If
individuals with FXS only showed impaired belief reason-
ing because they struggled to meet the incidental processing
demands of false belief tasks, then we would expect similar
levels of performance in false belief and comparison trials.
Another possibility was that the same cognitive resources
(e.g., working memory) that were necessary for handling the
incidental processing demands of false belief tasks were also
necessary for belief reasoning itself. If this were the case, we
might observe greater impairment on false belief trials than
on comparison trials (because false belief trials would make
greater overall demands on impaired cognitive resources),
but this impairment should be in proportion to the level
of impairment on comparison trials. Finally, if impairment
on false belief trials was disproportionate to impairment on
comparison trials then this would be evidence of a primary
deficit in “theory of mind” that was not merely the conse-
quence of impaired working memory.
Second, our video-based tasks allowed us to assess
whether individuals with FXS had particular difficulty resist-
ing interference from knowledge of reality when reasoning
about false beliefs. Resisting interference from knowledge
of reality is an inhibitory demand that is thought by many
to be a key problem in false belief reasoning (Carlson &
Moses, 2001), and this demand was present in the reality-
known false belief task (as well as in the standard false belief
tasks) but absent in the reality-unknown false belief task. If
individuals with FXS performed proportionately better on
false belief tasks when they did not need to resist interfer-
ence from knowledge of reality this would suggest that a
primary deficit with inhibitory control was responsible for at
least some of their “theory of mind” problems.
In accordance with the findings of Garner et al. (1999) and
Cornish et al. (2005) that males with FXS (without autism)
performed similarly to other children with intellectual dis-
ability on tests of theory of mind, it was predicted that boys
with FXS (without autism) will not have a specific theory of
mind deficit. It was predicted that their performance would
be related to the executive demands (i.e., working memory
and inhibitory control) of the task. However, on the basis
Springer
20 J Abnorm Child Psychol (2007) 35:17–28
of studies of children with autism that have reported a spe-
cific deficit in theory of mind (Yirmiya et al., 1998), it was
predicted that boys with FXS who scored above the clini-
cal cut-off for autism on a screening questionnaire, would
show theory of mind impairments and executive functioning
deficits.
Method
Participants
Three groups of children took part in the study: 15 boys with
fragile X syndrome (FXS) who displayed few autistic fea-
tures (full mutation (FMR-1) and fully methylated); 15 boys
with fragile X syndrome (FXS-A) with significant autistic
features (full mutation (FMR-1) and fully methylated); 15
boys with intellectual disability (ID) of unknown aetiology
who had no history of autism or FXS. Each participant was
assessed for autism using the Social Communication Ques-
tionnaire (SCQ: Berument, Rutter, Lord, Pickles, & Bailey,
1999), which is a measure that can be completed by parents
and can be done without the guidance of professionals. The
SCQ was designed as a screening measure for the Autism
Diagnostic Interview-Revised (ADI-R: Lord et al., 1994) and
has been shown to have high diagnostic validity. The authors
suggest a cut-off point for autism spectrum disorders of 15.
This score was found to differentiate individuals with autism
from other diagnoses, excluding individuals with intellectual
disability, with a specificity of .80 and a sensitivity of .96.
The same cut-off point was found to differentiate individu-
als with autism from individuals with intellectual disability
without autism, with a specificity of .67 and a sensitivity of
.96 (Berument et al., 1999). The screen is suitable for partic-
ipants above the chronological age of 4 years and a mental
age of 2 years.
Families of children with FXS were contacted through
the Fragile X Society. Consenting participants completed an
initial screening questionnaire that was designed to provide
information on their child’s developmental level and symp-
toms associated with autism (using the SCQ). The aim was
to recruit boys who were of sufficient verbal ability to partic-
ipate and to separate boys with FXS unlikely to have autism
from those who are likely to meet the diagnostic criteria for
autism. Boys who were rated by parents as verbal and able
to use simple sentences were included. Whilst SCQ suggests
that a score of 15 and above is a possible indicator of autism,
the standardisation data showed that the mean score for chil-
dren with autism was 24.2. Consequently, boys with FXS
were included in the FXS group if they scored 14 or below
on their total SCQ score. Boys with FXS with an SCQ score
of 25 or above were included in the FXS-A group. From 110
completed screening questionnaires, 73 were selected on the
basis of having sufficient developmental ability. A total of
33 boys who scored between 15 and 24 on the SCQ were
not included in the main study and a further 10 children who
were eligible for inclusion could not be tested due to a variety
of factors including geographical location and child illness.
Boys with intellectual disability were recruited from special
needs schools. Teachers were asked to nominate boys who
had no history of autism or FXS, were verbal and able to
use sentences and provided a match for the chronological
age of the boys with FXS. A total of 21 boys were assessed
on measures of verbal and non-verbal ability. Of this initial
sample of 21 boys, 15 boys were selected to complete the full
battery on the basis that they provided a good match for the
FXS group. All of the 15 boys selected scored 14 or below
on the SCQ.
The three clinical groups were pairwise matched on
Chronological age (CA) and verbal mental age (VMA) calcu-
lated using the long form of the British Picture Vocabulary
Scale (BPVS) (Dunn, Dunn, & Whetton, 1997). The first
edition of the test was used since the norms for this test are
suitable for children up to the age of eighteen years. This
test is designed to measure receptive vocabulary and does
not require any reading, speaking or writing, only simple
responses to picture cards. The child is presented with a se-
ries of plates containing 4 pictures and is asked to indicate
which is the picture of the word that is said. Reliability and
validity of the test are acceptable (see Dunn, Dunn, & Whet-
ton, 1997). It is widely acknowledged that theory of mind
development has a close connection with the development
of language (e.g., Astington & Jenkins, 1999; De Villiers
& Pyers, 2002; Slade & Ruffman, 2005; Smith, Apperly, &
White, 2003), though the relative importance of syntax and
semantics in this relationship remains controversial. Whilst
the BPVS is primarily a measure of receptive vocabulary it
has been shown to be a reliable estimate of general language
ability in children with autism (Jarrold, Boucher, & Russell,
1997).
The Coloured Progressive Matrices (Raven, 1965)was
used to assess non-verbal mental age. This test is designed
to measure a person’s ability to form perceptual relations
and reason by analogy independent of language. It consists
of 36 items arranged in three sets of 12 items each. Each
item contains a figure with a missing piece. Below the figure
are six alternative pieces to complete the figure, only one
of which is correct. This test is known to have high internal
consistency and high test-retest reliability (see Raven, 1965).
Verbal ability rather than non-verbal ability was chosen to
match the groups since verbal mental age has consistently
been related to theory of mind development in typically de-
veloping children and children with autism (Happ
´
e, 1995).
In order to be included in the study, participants had a verbal
mental age of 4 years or above. Table 1 summarises mean
CA, VMA, VIQ and NVMA for each group.
Springer
J Abnorm Child Psychol (2007) 35:17–28 21
False belief trials: Woman sees man place object in one box. In her absence the man swaps
the objects location. Participant has to predict where woman will look for the object when she
returns. Participants receive feedback by viewing where she actually looks.
Working m emory trials: Woman sees man place object in one box. In her presence the man
swaps the objects location. Participant has to predict where woman will look for the object.
Participants receive feedback by viewing where she actually looks.
Fig. 1 Reality known task (This figure is adapted and extended from Apperly, Samson, & Humphreys, 2005)
Table 1 Participant characteristics: mean chronological age (CA),
verbal mental age (VMA), verbal IQ & non-verbal mental age (NVMA)
in years and months (standard deviations are shown in brackets)
Group C A VMA VIQ NVMA
FXS (n =15) 13:8 (3:0) 6:11 (2.0) 53.9 (13.2) 6:0 (1.3)
FXS-A (n =15) 12:5 (2:6) 6:8 (2.7) 53.4 (12.8) 6:4 (1.9)
ID (n =15) 13:9 (2:9) 6:11 (1.8) 55.8 (13.8) 6:8 (1.10)
Note. ID: Intellectual disability group, FXS: fragile X group, FXS-A:
fragile X group with many autistic features.
Materials
The video-based tasks used videos of human actors and were
presented on a laptop computer using Windows Media Player
software. The location change task (also known as the Sally-
Anne task; Baron-Cohen, Leslie, & Frith, 1985) employed
two easily distinguishable play-mobile figures, a ball, a small
plastic box, and a small plastic cupboard. The deceptive box
task (Perner, Frith, Leslie, & Leekam, 1989) employed an
egg box and a pen.
Procedure
The boys with FXS were tested in their homes, with the ex-
ception of one boy who was tested at school. The group with
intellectual disability were contacted through their school
and participants were tested at school. Every effort was made
to ensure that testing conditions within the home and school
environments were similar. Each participant was individu-
ally tested in a quiet room, with no interruption and testing
was conducted at a table or desk. All the tasks were given
in one session, which typically lasted one-hour, inclusive
of regular breaks. All participants completed an assessment
of verbal and non-verbal ability and a battery of theory of
mind tests (two video-based tasks, one location change and
one deceptive box task). The order of administration of tasks
was BPVS, video-based false belief task (reality unknown
or reality known counter-balanced across participants), De-
ceptive box task, Standard location change task, Raven’s
Coloured Progressive Matrices and video-based false be-
lief task (reality unknown or reality known counter-balanced
across participants).
In the “reality known” version of the video-based task,
participants watched a series of short ( ∼45 s) videos in
which they were asked to predict where a character will
search for a hidden object (see Samson, Apperly, Kathirga-
manathan, & Humphreys, 2005). The task principles were
explained to the participant at the beginning of the task, and
comprehension was checked with a number of warm-up trials
on which corrective feedback was given. Video presentation
was controlled manually by the experimenter, enabling the
time allowed for responding and the rate of progress to the
next video to be adapted to the needs of the participant. Pre-
dicting where the woman would look on false belief trials
required the participant to follow and remember the trans-
fer of the object from one box to another, and to take ac-
count of the woman’s false belief while resisting interference
from their own knowledge of the object’s true location (see
Fig. 1). Predicting her search on memory comparison tri-
als also required the participant to follow and remember the
transfer of the object, but did not require belief reasoning,
since the woman’s belief was true. After responding to the
test question about where the woman will first look for a
hidden object, the participant received feedback by viewing
the end of the video clip where the woman opens a box. False
belief and memory comparison trials were mixed with two
anti-strategy filler trials, designed to minimise the possibil-
ity that the participant could learn to solve test trials using
superficial strategies. In the “woman-absent” filler trial, the
woman left the room and the object was first removed and
then replaced in its original box. Thus, a participant who
succeeded on false belief trials by pointing to the empty box
whenever Character 1 has left the room would fail this trial.
In the “woman-present” filler trial, the woman remained in
the room and the object was first removed from and then
Springer
22 J Abnorm Child Psychol (2007) 35:17–28
The participant does now know which box contains a hidden object –
their task is to work this out from a clue given by the woman.
False Belief Trial
The woman looks in the boxes. In her absence, the
man swaps the boxes. The woman indicates where she thinks the
object is. Since the participant has to infer that the woman has a false
belief in order to locate the object, knowledge of the object’s location
cannot interfere with this belief reasoning.
Working Memory Trial
The woman looks in the boxes and indicates
where she thinks the object is. In her absence, the man swaps the
boxes. The participant must remember the box indicated by the
woman and follow the transfer of the boxes.
Fig. 2 Reality unknown task (This figure is adapted and extended from Apperly, Samson, & Humphreys, 2005)
replaced in its original box. If a participant passed memory
comparison trials by always pointing to the opposite box
from the one in which the object was initially placed then
they would fail this trial. Thus, in total participants viewed
two false belief trials, two memory comparison trials, and
two trials of each anti-strategy filler trial. Trials of the same
kind were always presented consecutively, and the order of
trial presentation was counterbalanced across participants.
In the “reality unknown” false belief task, participants
again watched a series of short videos (from Apperly, Sam-
son, Chiavarino, & Humphreys, 2004). This time, although
they knew that there was an object in one of two identical
boxes, participants did not initially know which. The partici-
pant’s job was to locate the object, and on each trial a female
character in the video gave them a helpful clue by placing a
marker on top of the box she thought contained the object
(see Fig. 2). The principles of the task were explained and
practised in a warm-up phase at the beginning of the task. Lo-
cating the object in false belief trials required the participant
to take account of the movement of the target object when the
boxes were swapped and to take account of the woman’s false
belief when interpreting her clue about the object’s location
(see Fig. 2). However, unlike the reality-known false belief
task, participants did not know the object’s true location,
so did not have to resist interference from this knowledge
when reasoning about the woman’s false belief. Locating the
object in memory comparison trials also required the partici-
pant to take account of the woman’s clue (at a point when the
woman had a true belief) and also required the participant
to track the movement of the target object when the boxes
were swapped. However, the memory comparison trial did
not require reasoning about false beliefs. After pointing to
where they thought the object was located the participant re-
ceived feedback by viewing the end of the video clip where
the boxes are opened.
False belief and memory comparison trials were mixed
with two filler trials designed to check that participants un-
derstood the task and could not succeed on false belief trials
by adopting superficial strategies. On “inhibition” filler tri-
als the participant already knew the location of the object
when the woman (who had a false belief) returned to the
room and indicated the wrong box. Thus, to point correctly
participants had to resist any tendency to point to the box
just indicated by the woman. On “true belief” filler trials the
woman had a true belief and so gave a correct clue about
the object’s location. Since participants received feedback at
the end of each trial, these “true belief” trials provided feed-
back that should have prevented participants from adopting
the strategy of always pointing to the opposite box from the
one indicated by the woman. In total, participants viewed
two trials of each type. Trials of the same type were always
presented consecutively, and the order of trial presentation
was counterbalanced across participants.
Participants also completed one trial of the standard loca-
tion change task (Baron-Cohen et al., 1985) and one trial of
the deceptive box task (Perner et al., 1989). Details of these
tasks are presented in Appendix A.
Results
Participant characteristics
There were no significant differences between groups on
Chronological Age (F(2, 42) = 1.15, p = .32), Ver-
bal Mental Age (F(2, 42) = 0.05, p = .95), Verbal IQ
(F(2, 42) = 0.13, p = .88) or Nonverbal Mental Age
(F(2, 42) = 0.51, p = .61).
Comparison of performance on belief reasoning tasks
We initially analysed false belief scores without taking ac-
count of performance on control questions for standard false
belief tasks or comparison trials for the video-based tasks.
This analysis served two objectives. First, it provided a com-
prehensive test of whether individuals with FXS simply
Springer
J Abnorm Child Psychol (2007) 35:17–28 23
showed an overall impairment in belief reasoning in com-
parison with matched controls (regardless of whether this
impairment was primary, or the result of impairments to ex-
ecutive function or working memory). Second, this analysis
allowed us to assess whether the simple absence of the need to
resist interference from knowledge of reality (in the reality-
unknown video-based task) enabled children to reason more
successfully about false beliefs than when they also needed
to meet this incidental task demand (in the standard false be-
lief tasks and reality-known video-based task). Scores across
the two standard false belief tasks were summed to give
one Standard task score. This yielded group means of 0.60
(SD = .74), 0.40 (SD = .63) and 1.20 (SD = .77) for
the FXS, FXS-A and Intellectual Disability groups respec-
tively. The mean scores (out of two trials) for each group
on the reality known and reality unknown false belief tasks
are shown in Table 2. To examine whether performance on
false belief tasks was disproportionately impaired in children
with FXS, and whether performance varied across the three
types of false belief task, a mixed ANOVA with Group (FXS,
FXS-A, Intellectual Disability) as the between-participants
factor and performance on the three false belief tasks (Stan-
dard, reality-known, reality-unknown) as repeated measures
was conducted. There was a significant main effect of group,
F(2, 42) = 10.91, p < .0001. Post-hoc LSD comparisons
showed both FXS and FXS-A groups to have lower scores
than the intellectual disability group (p = .003, p < .001 re-
spectively), but scores for the FXS and FXS-A groups did not
differ from each other (p = .184). There was no significant
effect of task, F(2, 84) = .575, p = .565, and no signifi-
cant interaction between task and group, F(4, 84) = .263,
p = .901. Thus, both the FXS and FXS-A groups performed
worse on the three types of false belief tasks than the Intel-
lectual Disability control group.
False belief versus comparison trials on video-based tasks
To assess whether children in the sample had impairments
to belief reasoning that were disproportionate to any im-
pairment of executive function (e.g., working memory and
inhibitory control), we conducted further analyses taking
account of performance on comparison trials. Because the
comparison trial data and false belief trial data from the
video-based tasks came from independent trials we were
able to pursue the analytic strategy of assessing statistically
whether impairment on false belief trials was disproportion-
ate to impairment on comparison trials.
Table 2 shows mean scores by group on the video-based
reality known and reality unknown belief reasoning tasks.
The data were analysed using a mixed ANOVA with Group
(FXS, FXS-A, Intellectual Disability) as the between- partic-
ipants factor and trial type (false belief and working memory
comparison) as the within-participants factor.
Table 2 Mean scores by group on video-based reality unknown
and reality known false belief tasks (standard deviations are shown in
brackets)
Reality known Reality unknown
Working Working
Group False belief memory False belief memory
FXS (n =15) 0.40 (0.74) 1.87 (0.35) 0.73 (0.70) 1.20 (0.86)
FXS-A (n =15) 0.27 (0.59) 1.53 (0.64) 0.33 (0.62) 0.93 (0.88)
ID (n =15) 1.13 (0.92) 1.87 (0.35) 1.13 (0.92) 1.73 (0.46)
Note. ID: Intellectual disability group, FXS: fragile X group, FXS-A:
fragile X group with many autistic features.
For the reality known false belief task, there was a sig-
nificant main effect of group (F
(2,42)
= 5.32, p = .009,
η
p
2
= .20) and trial type (F
(1,42)
= 106.94, p < .001,
η
p
2
= .72), and a significant interaction (F
(2,42)
= 3.84,
p = .03, η
p
2
= .15). Simple effects analyses showed this
interaction to be due to differences between groups on false
belief trials, with both FXS and FXS-A performing worse
than the Intellectual Disability group. Thus, the FXS and
FXS-A groups performed worse than the Intellectual Dis-
ability control group on the false belief trials.
To check whether participants learned from the feedback
provided at the end of false belief trials we compared perfor-
mance on the first false belief trial with performance on the
second false belief trial. Performance did not change signif-
icantly from trial 1 to trial 2 for any group (all ps > .5 by
sign test).
We also checked performance on filler trials to assess
whether children showed a strong tendency for adopting in-
correct response strategies. On “woman present” filler trials
there were 4/30 errors in the FXS group, 5/30 errors in the
FXS-A group, and 4/30 errors in the ID group. On “woman
absent” filler trials there were 5/30 errors in the FXS group,
4/30 errors in the FXS-A group, and 0/30 errors in the ID
group. This indicated that participants were not generally
succeeding on test trials by adopting strategies linked super-
ficially to the presence or absence of the woman at the point
when the object was manipulated.
For the reality unknown false belief task, there was a
significant main effect of group (F
(2,42)
= 7.89, p = .01,
η
p
2
= .27)and trial type (F
(1,42)
= 12.98,p = .01, η
p
2
= .24).
There was no significant interaction (F
(2,42)
= 0.83,
p = .92, η
p
2
= .004). Pair-wise comparisons showed that
the FXS group differed from the Intellectual Disability group
( p = .026) but not from the FXS-A group ( p = .11). The
FXS-A group also differed from the Intellectual Disability
group (p = .001). Thus, the FXS and FXS-A groups per-
formed worse than the Intellectual Disability control group
on both false belief trials and working memory comparison
trials.
Springer
24 J Abnorm Child Psychol (2007) 35:17–28
Table 3 Number of
participants by group passing or
failing standard false belief
memory and test questions
FXS (n =15) FXS-A (n =15) ID (n =15)
Performance
Deceptive
box
Location
change
Deceptive
box
Location
change
Deceptive
box
Location
change
Passed FB, passed memory 6 3 3 3 9 9
Failed FB, passed memory 8 11 9 10 6 6
Failed FB, failed memory 1 1 3 2 0 0
Note. ID: Intellectual disability
group, FXS: fragile X group,
FXS-A: fragile X group with
many autistic features.
To check whether participants learned from the feedback
provided at the end of false belief trials we compared perfor-
mance on the first false belief trial with performance on the
second false belief trial. Performance did not change signif-
icantly from trial 1 to trial 2 for any group (all ps > .5 by
sign test).
We also checked performance on filler trials to assess
whether children showed a strong tendency for adopting in-
correct response strategies. On true belief filler trials there
were 4/30 errors in the FXS group, 3/30 errors in the FXS-A
group, and 5/30 errors in the ID group. This pattern confirmed
that participants understood the meaning of the woman’s clue
and were not achieving correct responses on false belief tri-
als by pursuing a general strategy of pointing to the opposite
box from that indicated by the woman. On “inhibition” filler
trials there were 4/30 errors in the FXS group, 5/30 errors
in the FXS-A group, and 3/30 errors in the ID group. This
indicated that participants were not generally failing on false
belief trials merely because they always pointed to the box
indicated by the woman.
Standard false belief tasks
Since the control questions of the Standard false belief tasks
were asked within the same trial as the false belief question
itself, failure on the control questions may have directly
affectedresponses on the false belief question. Consequently,
we did not include data from children who answered control
questions incorrectly.
Performance on the two individual standard false belief
tasks was analysed and the number of participants by group
passing these tasks is shown in Table 3. A small number
of children made errors on the memory check questions of
the location change and deceptive box false belief tasks.
Data from these children were excluded for the following
analyses.
On the location change task, the FXS and FXS-A groups
both performed poorly compared to the Intellectual Disabil-
ity group (χ
2
= 4.441, df = 1, p = .035; χ
2
= 3.877,
df = 1, p = .049 respectively). Performance of the FXS,
and FXS-A groups did not differ (Fisher’s exact = 0.11,
df = 1, p = .638). On the deceptive box task, there were
no significant differences between groups passing/failing the
false belief test question (χ
2
= 3.325, df = 2, p = .19).
Thus, after excluding participants who failed the memory
control questions, both the FXS and FXS-A groups per-
formed worse than the Intellectual Disability control group
on the location change task, but there were no statistically
significant differences between groups on the deceptive box
task.
Discussion
Our first question was whether children with FXS would be
disproportionately impaired on theory of mind tasks (such
as false belief tasks). Our findings suggest that children with
FXS do have a clear impairment on false belief tasks that
is disproportionate to their general level of intellectual dis-
ability. The absence of learning from feedback on our tasks
indicated that this difficulty is robust, and not the result of
a simple misunderstanding of the point of the tasks. Impor-
tantly, this was the case whether or not the children with
FXS had significant autistic features, suggesting that theory
of mind difficulties in FXS are not merely an artifact of the
high co-morbidity of FXS and autism. Indeed, while there
was a numerical trend for the FXS group to make fewer er-
rors on false belief tasks than the FXS-A group, this was
non-significant for the sample size reported here. It is possi-
ble that differences between the FXS and the FXS-A groups
would emerge using a larger sample size. This remains an
important area of investigation for future studies.
The findings of the present study are broadly in line with
those reported by Garner et al. (1999), whose findings sug-
gested that boys with FXS performed poorly on tests of first
order false belief compared with mental age matched con-
trols. Fewer boys with FXS passed the deceptive box task
and the location change task than the control group, although
this difference only reached statistical significance on the de-
ceptive box task. The FXS group reported in Cornish et al.
(2005) performed better on tests of first order false belief
than the FXS groups reported in our study and the FXS
group reported in Garner et al. (1999). Just under half the
FXS group passed the location change task in Cornish et al.
(2005), which was similar to the pass rate for the compari-
son group of children with Down syndrome included in the
study. It is unclear why the FXS group in the Cornish et al.
(2005) performed comparably to the control group whilst in
the present study and the Garner study the FXS groups were
out-performed by controls. One possibility is that the control
Springer
J Abnorm Child Psychol (2007) 35:17–28 25
group in the Cornish et al. (2005) had an unusually poor per-
formance. A meta-analysis of some 40 studies by Yirmiya,
Erel, Shaked, and Solomonica-Levi (1998) has highlighted
that people with Down syndrome show a unique strength on
tests of theory of mind. It is therefore puzzling why less than
half of the group with Down syndrome passed the first order
false belief tests despite a relatively high verbal mental age
of 6 years 7 months.
Our second question was whether any theory of mind
impairment was primary, or a consequence of impaired in-
hibitory control or working memory. We noted that standard
false belief tasks confound the need to reason about beliefs
(e.g., Simon thinks the chocolate is in the cupboard) with
the need to resist interference from one’s own knowledge
of the correct answer (that in fact the chocolate is in the
fridge). Since difficulty with executive processes such as re-
sisting interference is known to be a distinctive feature of
the FXS cognitive phenotype (Cornish, Sudhalter, & Turk,
2004), children with FXS might have made errors on such
false belief tasks merely because they failed to meet this ex-
ecutive demand. This hypothesis was tested by comparing
performance on traditional “reality-known” false belief tasks
with performance on the relatively novel “reality-unknown”
false belief tasks, where this executive confound is removed.
The results showed that, compared with the Intellectual Dis-
ability control group, children with FXS made more errors on
all types of false belief task (although on the standard tasks
this only reaches statistical significance when the tasks are
combined), and that neither the Intellectual Disability nor
FXS children found the reality-unknown false belief task
any easier than reality-known false belief tasks. This result
suggests that difficulty resisting interference from knowl-
edge of reality could not explain difficulty on the false belief
tasks for any group. It is also noteworthy that the standard
false belief tasks required children to comprehend verbal
narratives acted out with puppets, whereas the video-based
false belief tasks made relatively few demands on verbal
comprehension. The absence of any difference in perfor-
mance across these types of task suggests that the verbal
comprehension demands of standard false belief tasks may
not be a significant source of difficulty for children with
FXS.
It was also noted that false belief tasks make substan-
tial demands on working memory. Since impaired working
memory is known to be a feature of the FXS cognitive pheno-
type, it is important to test whether impaired working mem-
ory alone could explain any errors made on false belief tasks.
The commonly used location change and deceptive box false
belief tasks only allow relatively weak checks to be made,
by testing whether children remember crucial facts about the
story. When data from children who failed these memory
checks were discounted, children with FXS still performed
less well than the ID control group on the location change
false belief task. However, for the deceptive box task, once
data from children who failed the check question were dis-
counted, differences between the groups’ performance were
not significant.
The video-based reality-known and reality-unknown false
belief task allowed a stronger test of the role of working
memory because they incorporated independent compari-
son trials. For the reality-known task, the comparison trial
was analogous to the “reality” control question of the lo-
cation change false belief task, and merely tested whether
the participant had followed the visible displacement of the
object from one container to the other. Compared with In-
tellectual Disability controls, children with FXS were still
disproportionately impaired on false belief trials, even when
these memory comparison trials were taken into account.
Memory comparison trials for the reality-unknown task were
more demanding, and more closely matched to the incidental
memory demands of the reality-unknown false belief trials.
For both false belief and comparison trials, participants did
not see the critical hidden object, but had to realize that it
had changed locations when boxes were swapped. Analy-
ses showed that children with FXS had significantly more
difficulty than Intellectual Disability controls on both the
false belief and the memory comparison trials. That is to
say, the difficulty that children with FXS experienced on
false belief trials was proportionate to this group’s impair-
ment on working memory trials. This pattern of results can
be interpreted in two ways. One possibility is that children
with FXS have independent impairments of theory of mind
and working memory. The second possibility is that chil-
dren with FXS made errors on false belief trials because of
the substantial demands that belief reasoning makes on their
impaired working memory processes, not because of inde-
pendent impairment to a neuro-cognitive system devoted to
belief reasoning. Evidence to discriminate between these al-
ternatives might come from assessing individual differences
in working memory and theory of mind in a much larger
sample of children with FXS. However, in the absence of
such evidence, we favour the more parsimonious account
that attributes theory of mind impairment in children with
FXS to a primary deficit in working memory.
The present study employed participants recruited from
the Fragile X Society and schools, rather than from a clini-
cal setting. Recruiting participants from these sources rather
than recruiting from a clinic was undertaken to ensure the
study employed a representative sample. In this respect, em-
ploying a non-clinical sample was a strength of the study.
However, employing a non-clinical sample also meant that
the FXS children were divided into FXS and FXS-A groups
using the SCQ (Berument et al., 1999), a parent report mea-
sure of symptoms associated with autism rather than being
divided into groups using a clinical diagnosis of autism.
The SCQ is an autism screening measure derived from the
Springer
26 J Abnorm Child Psychol (2007) 35:17–28
Autism Diagnostic Interview-Revised (ADI-R: Lord et al.,
1994), which is one of the most widely used instruments in
the diagnosis of autism. A recent study by Howlin and Karpf
(2004) used the SCQ to identify autism spectrum disorders
in Cohen Syndrome. The findings suggested that while the
SCQ cut-off score of 15 and above outlined by Berument
et al. (1999) was successful at identifying cases of autism
spectrum disorder, there were some false negatives (indi-
viduals who did not meet the criteria on the SCQ but met
criteria for autism spectrum disorders on the ADI-R). The
implication of these findings for the present study is that
the individuals in the FXS-A group are likely to meet the
diagnostic criteria for autism on the ADI-R. However, it is
possible that some of the individuals in the FXS group would
also meet the diagnostic criteria for autism on the ADI-R. A
further issue is that in forming the FXS groups, we excluded
boys who had scores on the SCQ between 15 and 24. This
group was excluded in an attempt to reduce the likelihood
of false positives for autism in the FXS-A group. Conse-
quently, we cannot conclude from our data how this group
would perform on theory of mind measures. These method-
ological limitations should be addressed in future research.
Whilst acknowledging that there are only a handful of
studies of theory of mind in boys with FXS, there are some
issues identified that may be of value to consider in a clinical
setting and to explore in future research. There is evidence
that theory of mind is impaired in boys with FXS and it is
possible that this deficit plays a role in the social difficulties
associated with FXS. In particular, males with FXS experi-
ence difficulties with social interaction and communication,
development of which is positively associated with theory of
mind ability in typically developing children (e.g., Astington,
2000). Interventions aimed at improving theory of mind un-
derstanding in young children with FXS may therefore help
to improve the developmental outcomes of these children.
Work undertaken to improve theory of mind in children with
autism has found that thought-bubbles are a successful strat-
egy for developing mental-state understanding (Wellman,
Baron-Cohen, Caswell, Gomez, Swettenham, Toye et al.,
2002). The authors cautioned that whilst this technique did
enable the majority of the children with autism to pass tests
of false-belief and to generalize their understanding to other
tests, some children learnt very little. Moreover, the gener-
alizability of this training to real-life social situations has
been disappointing (Hadwin, Baron-Cohen, Howlin, & Hill,
1997).
Overall, the findings of the research indicate that boys
with FXS have difficulty with theory of mind tasks com-
pared to an intellectual disability control group. However,
boys with FXS also perform worse on comparison trials.
This means that theory of mind difficulties are likely to be
an important aspect of the FXS clinical profile. However, it
is probable that the deficits in theory of mind understand-
ing stem from general information processing deficits (e.g.,
working memory), not from a primary deficit in theory of
mind, and that this is the case whether or not children with
FXS have significant autistic features.
Acknowledgements We would like to thank Lynne Zwink for her
invaluable support, the families recruited through the Fragile X Society
and the staff and pupils at Foxwood School, who generously gave their
time to take part in this research.
Appendix A
Standard location change false belief task. Based upon the
Sally-Anne task (Baron-Cohen et al., 1985), the participant
is introduced to two play-mobile figures and is invited to
give them names (for example James and Carol). The ex-
perimenter then narrates the following story: “This is James
and this is Carol. James has just come into the house from
playing football outside. He takes his ball and puts in into
his toy box and goes to play upstairs”. The participant sees
the play-mobile figure put his ball into a box and exit the
area. The narrative is then continued. “While James is play-
ing upstairs, Carol decides to move the ball. Carol takes the
ball out of the box and puts it in the cupboard”. The par-
ticipant sees the second play-mobile figure remove the ball
from the box and put it into the cupboard, where it is out of
sight. This second figure then exits the area. The participant
is then asked two control questions about the original and the
current location of the ball: Memory Question: “Where did
James put his ball in the beginning?” and the Reality Ques-
tion “Where is the ball now?” The first play-mobile character
is then brought back into the area, and the participant is told
“James is going to play outside again now and he wants
his ball”. Then the participant is asked the Test Question,
“Where will James look for his ball?” The participant scores
one for a correct response or zero for an incorrect response
on the test question.
Standard Deceptive box false belief task. In this variant
of the deceptive box task (Perner, Frith, Leslie, & Leekam,
1989), the participant is shown a sealed egg carton and asked,
“What do you think is inside here?” The participant answers
“eggs” or “those” (pointing to the picture of the eggs on the
box). The box is then opened up and the participant is shown
that the box actually contains a pen. The participant is asked
to name the pen to check that they know what it is. The pen
is put back in the box and the box is closed. The participant
is then asked, “In a minute X is going to come and work with
me. (S)he hasn’t seen this box yet, or what’s inside it. When
(s)he comes in, I am going to show her/him this box, closed
just like this (participant is shown box again). I am going to
ask her/him “X, What’s in this box?” The participant is then
asked the Test Question “What will (s)he say?” and a Reality
Question, “What is really inside the box?” The participant
Springer
J Abnorm Child Psychol (2007) 35:17–28 27
scores one for a correct response or zero for an incorrect
response on the test question.
References
Apperly, I. A., Samson, D., Chiavarino, C., & Humphreys, G. W. (2004).
Frontal and temporo-parietal lobe contribution to theory of mind:
Neuropsychological evidence from a false belief task with reduced
language and executive demands. Journal of Cognitive Neuro-
science, 16(10), 1773–1784.
Apperly, I. A., Samson, D., & Humphreys, G. W. (2005). Domain-
specificity and theory of mind: Evaluating evidence from neu-
ropsychology. Trends in Cognitive Sciences, 9(12), 572–577.
Astington, J. W. (2000). Minds in the Making: Essays in honour of
David R. Olson. Oxford: Blackwell.
Astington, J. W., & Jenkins, J. M. (1999) A longitudinal study of
the relation between language and theory-of-mind development.
Developmental Psychology, 35, 1311–1320.
Bailey, D. B., Mesibov, G., Hatton, D. D., Clark, R. D., Roberts, J. E., &
Mayhew, L. (1998). Autistic behavior in young boys with fragile-
X syndrome. Journal of Autism and Developmental Disorders, 28,
499–508.
Bailey, D. B., & Nelson, D. (1995). The nature and consequences
of fragile X syndrome. Mental Retardation and Developmental
Disability Research Review, 1, 238–244.
Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic
child have a theory of mind? Cognition, 21, 37–46.
Baron-Cohen, S. (2000). Theory of mind and autism: A fifteen year
review. In S. Baron-Cohen, H. Tager-Flusberg, & D. J. Cohen
(Eds.), Understanding other minds: Perspectives from develop-
mental cognitive neuroscience. Oxford: Oxford University Press.
Belser, R. C., Sudhalter, V. (2001). Fragile X syndrome: repetitive
speech. American Journal of Mental Retardation, 106, 28–38.
Bennetto, L., & Pennington, B. F. (2002). Neuropsychology. In R. J.
Hagerman & P. J. Hagerman (Eds.), Fragile X syndrome: Diag-
nosis, treatment, and research (3rd ed., pp. 206–248). Baltimore,
MD: Johns Hopkins University Press.
Bennetto, L., Pennington, B. F., Porter, D., Taylor, A. K., & Hagerman,
R. J. (2001). Profile of cognitive functioning in women with the
fragile X mutation. Neuropsychology, 15, 290–299.
Berument, S. K., Rutter, M., Lord, C., Pickles, A., & Bailey, A. (1999).
Autism screening questionnaire: Diagnostic validity. British Jour-
nal of Psychiatry, 175, 444–451.
Carlson, S. M., & Moses, L. J. (2001). Individual differences in in-
hibitory control and children’s theory of mind. Child Development,
72, 1032–1053.
Cornish, K., Burack, J., Rahman, A., Munir, F., Russo, N., & Grant,
C. (2005). Theory of mind deficits in children with Fragile X
syndrome. Journal of Intellectual Disability Research, 49, 372–
378.
Cornish, K. M., Munir, F., & Cross, G. (2001). Differential impact of the
FMR-1 full mutation on memory and attention functioning: A neu-
ropsychological perspective. Journal of Cognitive Neuroscience,
13, 144–150.
Cornish, K., Sudhalter, V., & Turk, J. (2004). Attention and language
in fragile X. Mental Retardation and Developmental Disabilities
Research Reviews, 10, 11–16.
de Villiers, J. G., & Pyers, J. E. (2002) Complements to cognition: A
longitudinal study of the relationship between complex syntax and
false-belief- understanding. Cognitive Development, 17, 1037–
1060.
Dunn, L. M., Dunn, L. M., Whetton, C., & Burley, J. (1997). British
picture vocabulary scale (2nd ed.) (BPVS-II). Windsor, Berks:
NFER-Nelson.
Dykens, E. M., & Volkmar, F. R. (1997). Medical conditions associated
with autism. In D. J. Cohen & F. R. Volkmar (Eds.), Handbook
of autism and pervasive developmental disorders (2nd ed.). New
Yo rk : Wi ley.
Frith, U., & Frith, C. D. (2003) Development and neurophys-
iology of mentalizing. Philosophical Transactions of the
Royal Society of London B. Biological Sciences, 358, 459–
473.
Garner, C., Callias, M., & Turk, J. (1999). Executive function and theory
of mind performance of boys with fragile-X syndrome. Journal of
Intellectual Disability Research, 43, 466–474.
Hadwin, J., Baron-Cohen, S., Howlin, P., & Hill, K. (1997). Does
teaching theory of mind have an effect on the ability to develop
conversational ability in children with autism? Journal of autism
and Developmental Disorders, 27, 519–537.
Happ
´
e, F. G. E. (1995). The role of age and verbal ability in the theory
of mind task performance of subjects with autism. Child Develop-
ment, 66, 843–845.
Howlin, P., & Karpf, J. (2004). Using the social communication ques-
tionnaire to identify ‘autistic spectrum’ disorders associated with
other genetic conditions. Autism, 8, 175–182.
Jarrold, C., Boucher, J., & Russell, J. (1997). Language profiles in
children with autism. Autism, 1, 57–76.
Loesch, D. Z., Bui, Q. M., Grigsby, J., Butler, E., Epstein, J., Huggins,
R. M., Taylor, A. K., & Hagerman, R. J. (2003). Effect of fragile
X status categories and the fragile X mental retardation protein
levels on executive functioning in males and females with fragile
X. Neuropsychology, 17, 646–657.
Lord, C., Rutter, M., & Le Couteur, A. (1994). Autism diagnostic
interview-revised: A revised version of a diagnostic interview
for carers of individuals with possible pervasive developmental
disorders. Journal of Autism and Developmental Disorders, 24,
659–685.
Mazzocco, M. M. M., Pennington, B. F., & Hagerman, R. J. (1994).
Social cognition skills among females with fragile X. Journal of
Autism and Developmental Disorders, 24, 473–485.
Ozonoff, S., Pennington, B. F., & Rogers, S. J. (1991). Executive func-
tion deficits in high-functioning autistic individuals: Relationship
to theory of mind. Journal of Child Psychology and Psychiatry,
32, 1081–1105.
Perner, J., Frith, U., Leslie, A. M., & Leekam, S. (1989). Exploration
of the autistic child’s theory of mind: Knowledge, belief and com-
munication. Child Development, 60, 689–700.
Perner, J., & Lang, B. (2000). Theory of mind and executive func-
tion: Is there a developmental relationship. In S. Baron-Cohen, T.
Tager-Flusberg, & D. Cohen (Eds.), Understanding other minds:
Perspectives fromdevelopmentalcognitiveneuroscience (pp. 150–
181). Oxford: Oxford University Press.
Raven, J. C. (1965). Guide to using the coloured progressive matrices.
London: Lewis.
Russell, J. (1997). How executive disorders can bring about an inade-
quate theory of mind. In J. Russell (Ed.), Autism as an executive
disorder (pp. 256–304). Oxford: Oxford University Press.
Samson, D., Apperly, I. A., Kathirgamanathan, U., & Humphreys, G. W.
(2005). Seeing it my way: A case of selective deficit in inhibiting
self-perspective. Brain, 128, 1102–1111.
Saxe, R., Carey, S., & Kanwisher, N. (2004). Understanding other
minds: Linking developmental psychology and functional neu-
roimaging. Annual Review of Psychology, 55, 87–124.
Slade, L., & Ruffman, T. (2005). How language does (and does not)
relate to theory of mind: A longitudinal study of syntax, semantics,
working memory and false belief. British Journal of Developmen-
tal Psychology, 23, 1–26.
Smith, M., Apperly, I. A., & White, V. (2003) False belief reasoning and
the acquisition of relative clause sentences. Child Development,
74(6), 1709–1719.
Springer
28 J Abnorm Child Psychol (2007) 35:17–28
Sperber, D. (Ed.). (2000). Metarepresentations: A multidisciplinary
perspective. New York: Oxford University Press.
Sudhalter, V., & Belser, R. C. (2001). Conversational characteristics of
children with fragile X syndrome: tangential language. American
Journal of Medical Genetics,Sep;106(5), 389–400.
Wellman, H. M., Baron-Cohen, S., Caswell, R., Gomez, J. C., Swetten-
ham, J., Toye, E., & Lagattuta, K. (2002). Thought-bubbles help
children with autism acquire an alternative to a theory of mind.
Autism, 6, 343–363.
Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: repre-
sentation and the constraining function of wrong beliefs in
young children’s understanding of deception. Cognition, 13, 103–
128.
Yirmiya, N., Erel, O., Shaked, M., & Solomonica-Levi, D. (1998).
Meta-analysis comparing theory of mind abilities of individu-
als with autism, individuals with mental retardation, and nor-
mally developing individuals. Psychological Bulletin, 124, 283–
307.
Springer