Baron-Cohen et al. (2001)

Key Notes
Autism Spectrum Disorder (ASD) = A neurodevelopmental disorder impairing a child's ability to communicate and interact. Symptoms: repetitive behaviour, little or no eye contact and can't recognise facial expressions. High Functioning Autism is HFA.

Aspergers Syndrome (AS) = Autism spectrum disorder that affects language and communication skills. Symptoms: restricted & repetitive behaviours and trouble identifying facial expressions.

Autism Spectrum Quotient Test (AQ) = self-report questionnaire with scores ranging from 0 to 50. A high score suggests that the person has more autistic traits.

Theory of Mind (ToM) = the ability to understand the view of another.

The original study (1997)
The original study had numerous issues. Participants were presented with 25 photos showing different eye expressions and they chose the mental state shown between 2 options.
Issues with the original study:
1. It had more female faces than male faces. Solution: equal no. of male and female faces in the RET question.
2. Test had both basic and complex mental states. The basic ones were too easy. Solution: Only complex mental states were used.

Aim
1. To test if the revised version of the 'Reading the Mind in the Eyes' test would be successful at differentiating participants with AS or HFA (High Functioning Autism) from the general population.
2. To test if there is a negative correlation in a sample of normal adults and between the RET and the AQ.
3. To test any sex differences on the RET when normal adults take the test.

Hypotheses
1. Participants with autism will score significantly lower in the RET than the control group.
2. Participants with autism will score significantly higher on the AQ test.
3. Females in the 'normal' group (grp 2 & 3) will score higher on the RET than males in those groups.
4. Males in the 'normal' group will score higher on the AQ measure than females.
5. Scores on the AQ and RET will be negatively correlated.

Background
Baron-Cohen suggested that people with autism have an undeveloped 'ToM'. To measure the 'ToM' of individuals, Baron-Cohen came up with the 'Reading the Mind in the Eyes' task where participants would be evaluated on their ability to label others' emotions by observing their eye expressions on photos.

Sample
Group 1: 15 male adults with AS/HFA. Self-selecting sampling from the UK National Autistic Society through a magazine advert. Average IQ of 115; from a mix of socio-economic and educational backgrounds.
Group 2 - Comparative control group of adults: 122 adults from the adult community & educational classes from Exeter and public library users in Cambridge. 55 males and 67 females. From a broad mix of occupations and education.
Group 3 - Comparative control group of students: 103 undergraduate students from Cambridge. 53 males and 50 females. Assumed to have IQ higher than other participants.
Group 4 - IQ matched controls: 14 randomly selected adults whose IQ matched with that of group 1. Average IQ of 116.

Research Method, Design and Variables
Quasi-experiment.
Independent groups design
Independent Variables: Whether they had AS/HFA or were normal, and gender.
Dependent Variables: RET score, AQ score, and gender identification for group 1.

Procedure
A. Developing the Revised Eye Test (RET): Baron-Cohen and Wheelwright originated target words and foils for 36 photos. It was piloted on 8 judges (4 males and 4 females). At least 5 judges had to agree on the target word and no more than 2 judges could select any single foil word. If this did not happen, the target word and or foil would be repiloted until the criteria were met for each.

In pilot tests, groups 2 and 3 achieved 100% on judging gender. The control group were tested with 40 photos but 4 were eliminated, resulting in 36 items. When results were being calculated, only the data for the 36 eye sets were taken.

B. Implementing the Tests: Each test was individually administered in a quiet room at either Cambridge or Exeter. There was no time limit. Each participant was given a practice test and then presented with 36 sets of eyes and 4 possible target words. Group 1 judged the gender of each photo as a control task. Participants in groups 1, 3 & 4 completed the AQ test. They read through the glossary of terms and were asked to ask questions as needed; they could use the glossary during the test.

Ethics
Participants' consent was taken and they knew the nature of the study. Group 1 participants had been diagnosed in specialist centres using APA criteria. The data collected was anonymised.

Results
• Scores ranged from 17 - 35, with a mode of 24.
• Adults with AS/HFA performed significantly worse than other groups for the RET (H1 supported).
• On the AQ test, adults with AS/HFA performed significantly higher than the control groups (H2 supported).
• Females scored higher on the RET (H3 supported).
• There was a significant negative correlation (-0.53) between scores on AQ and RET (H5 supported).

Conclusions
• Current study replicated findings that AS/HFA adults are significantly impaired in identifying the emotions of others.
• Current study replicated findings that AS/HFA adults score significantly higher on the AQ test than the general population.
• There were gender differences found on the RET as females performed better than men. However, it would have been more significant with a greater sample.
The Revised Eye Test was a more sensitive measure of adult social intelligence.

Strengths and Weaknesses
S1:It was a lab experiment, so confounding variables could be controlled. Thus, there is internal validity and it is easily replicable. Everyone saw the same set of eyes.
S2:Improvements on the eyes test improved validity.
W1:The study lacks ecological validity as eye expressions in real life are quick, and not static.
W2:The experimental sample (group 1) is small, so generalising results to those with AS/HFA is not possible.
W3:Only the eyes were used, but normally we study the whole face (mouths are expressive) therefore it lacks ecological validity.
W4:As this was a quasi-experiment, it wasn't possible to randomly allocate participants to the conditions. This introduces a confounding variable as it could be another factor causing the defence in scores between groups. Researchers try to attend to this by having the IQ matched control group.

Application to Everyday Life
Programmes could be conducted to help people with AS/HFA in developing their skills of interpreting emotions. The eyes test could be improved to help diagnose individuals who may have underlying autistic disorders.

Individual vs. Situational Explanations
The AS/HFA group performed significantly worse on the RET than the 'normal' group. This suggests that the ability to identify mental states is an individual skill that is developed. The environment had been standardised (supports individual explanation).

Children as Participants
Sophisticated words would not be appropriate for children. So, the 'Anne and Sally' test was developed to understand the theory of mind of children.