Click Here For Latest Update Please

Does the autistic child have a 'theory of mind'?

Baron-Cohen, S., Leslie, A.M. & Frith, U. (1985).


This paper concerns an experimental approach, investigating the nature of autism. Autism manifests as the inability to communicate with others, and an obsession with a restricted repertoire of activities. Children are very much in their own little world, and show no regard for the existence of others.

Early explanations and approaches.

1.      Bettelheim (1967) poor parenting (Psychoanalytic)

2.      Tinbergen & Tinbergen (1983) interactions within families (Ethological)

3.      Lovaas (1979) suggested that the symptoms are dealt with, without going into the causes (Behaviourism).

Baron-Cohen et al's explanation

Baron-Cohen et al's approach is social-cognitive.

Autistic people do not have a 'Theory of Mind' (Premack & Woodruff, 1978). They do not understand that people have their own thoughts about the world. This has serious implications for social interaction. Perhaps you can think of what these implications are?

Details of the study

The Sally-Anne test was used (Wimmer & Perner, 1983). This test is designed to test for a child's ability to understand what a person believes, and can be easily solved by 'normal' five year olds.

Subjects

Participants

Type of participant

Mean Chronological Age

Mean Verbal Mental age

Autistic

11 years 11 months

5 years 5 months

Down's syndrome

10 years 11 months

2 years 11 months

'Normal' children

4 years 5 months

You work it out!

The Down's syndrome children are used as a control for intelligence. It might be argued that the autistic children are unable to pass the Sally-Anne test because they are of low intelligence. (In fact they are of lower intelligence than 'normal' children , most likely because of their poor communication skills, rather than an innate deficit). The Down's Syndrome children have a lower mean intelligence than the autistic sample's mean; So if the Down' s Syndrome children are able to pass the Sally-Anne test and the Autistic children can not, then the inability to pass the Sally-Anne test can not be owing to the Autistic children's level of intelligence.

The 'Normal' 4 year old children are a control for the effects of maturation. It could be argued that autistic children are unable to pass the test because they are not mature enough. If 4 year olds (note, with a lower mental age than the autistic children) can pass the test and eleven year old Autistic children can not, then failure to get the Sally-Anne test correct can not be due to the Autistic children's level of maturity.

Design

Independent Variable - Quasi-experimental design - the three groups of children - Autistic, Down's Syndrome, and (mean) four year olds.

Dependent Variable - success or failure with the Sally-Anne test

Procedure

Children tested individually.

Seated at a desk opposite to an experimenter.

Two dolls, Sally and Anne.

Sally has a basket in front of her and Anne has a box.

Naming Question - children, after being told the names of the dolls, are asked to confirm that they know the names of the two dolls. When you are familiar with the experiment you should consider why the 'naming' question is important.

Sally places a marble in her basket.

Sally goes for a walk (disappears from view).

Whilst Sally has gone, Anne plays a trick and takes the marble from Sally's basket and transfers it to her (Anne's) box.

Sally returns.

The child is asked the main experimental question (the 'belief' question).
"Where will Sally look for her marble?"
Correct response being "in Sally's basket", because that is where Sally left it, and she is unaware of Anne's trickery.
Incorrect response being "in Anne's box", because Sally does not know this and the child is just telling the experimenter where the marble really is (where they believe the marble is). This would demonstrate an inability to consider what Sally's beliefs are. Another incorrect response could be pointing to somewhere other than the basket or box, which could indicate that the child was not paying attention or completely mis-understands the task (in fact, this did not happen).

Two control questions are then asked - Reality Question and the Memory Question .

The 'Reality' question is "Where is the marble really?". This is to make sure the child had paid attention to the transfer of the marble from the basket to the box.

The 'Memory' question is "Where was the marble in the beginning?". This is designed to make sure the child had not forgotten where Sally had left her marble.

The test is repeated, but this time the marble is transferred by Anne to the experimenter's pocket.

Results.

Percentage of correct answers

Question

Autistic

Down's Syndrome

'Normal'

Naming

100

100

100

Reality

100

100

100

Memory

100

100

100

Belief

20

86

85

These results are significant at the level of p is less than 0.001.

Discussion.

All of the control questions (Naming, reality and memory) were answered correctly. The belief question was answered correctly by 20% of the autistic children, compared to 86% of the Down's Syndrome children and 85% of the four year olds. This suggests the autistic children could not appreciate what Sally believed. They lack a theory of mind.

The autistic children always pointed to where the marble really was; either in the box or in the experimenter's pocket

Autistic children are known not to engage themselves in pretend play, so the use of dolls rather than real people could have been a mistake. The study was replicated by Leslie and Frith (1988) using people rather than dolls, and the results followed the same pattern as the original experiment.

Click here for more

Relationship between finger size and autism


Your questions

1. Simon Baron-Cohen, Alan M.Leslie and Frith: Autism study

I was struggling to find 4 issues (not ethical) which the study raises.

The study is a relatively simple laboratory experiment, which supports the 'Theory of mind'.  This finding is useful, as looking for the absence of a ToM is a clear indicator in diagnosing autism as well as suggesting an appropriate treatment.

As with all studies there are methodological issues:

 

Return to Gary Sturt's homepage


Gary Sturt
gary.sturt@ntlworld.com

Copyright © 2000 Gary Sturt
This Home Page was created by WebEdit,
Most recent revision 22nd March 2001