Baron - Cohen

Reading the Mind in the Eyes (2001)

Cognitive Approach

studies the theory of mind - ability to understand mental/emotional states of others and understand that they're different from your own

Psychology

studies autism and theory of mind, and if there is a valid test to identify autism

Sally Anne test

(Background)

tests ability of young children to understand that not all people are good

Baren - Cohen's 1997 eye test

(Background)

used to identify adults with AS/HFA

AQ test

(Background)

50 questions and a 4-point scale, score of 32 or higher indicates significant level of autistic traits

1997 test issue #1

(Background)

two word choices, giving 50% chance of guessing correctly (fixed by four word choices)

1997 test issue #2

(Background)

foil and semantic words were exact opposites of each other (fixed by foils not being opposite of semantic)

1997 test issue #3

(Background)

more female faces than male faces (fixed by making them equal)

1997 test issue #4

(Background)

parents of children with AS/HFA scored the same level as the kids themselves (fixed by making the test more difficult by adding word choices and items)

1997 test issue #5

(Background)

ceiling effect - too many people scored too high on the test (fixed by making the test more difficult by adding word choices and items)

1997 test issue #6

(Background)

gazing - some times were guessed from gaze (fixed by removing cards where the person was gazing)

1997 test issue #7

(Background)

comprehension - not understanding words (fixed by giving participants a glossary)

Aim #1

to test whether the revised eye test will fix the problems of the 1997 test and if it can identify AS/HFA

Aim #2

to see if females will score better on the eyes test than males

Aim #3

to see if a negative correlation exists between the AQ test and eyes test

Hypothesis #1

The AS/HFA group will score lower on the eyes test

Hypothesis #2

The AS/HFA group will score higher on the AQ test

Hypothesis #3

Females will score higher on the eyes test than males

Hypothesis #4

Males will score higher than females on the AQ test

Hypothesis #5

Scores on the eyes test and the AQ test will be negatively/inversely correlated

IV

groups of people: AS/HFA (experimental), normal females, normal males, neuronormal (control)

DV

scores on the AQ test and eyes test

Method

- Quasi study - laboratory experiment but IV is naturally occurring so independent measures must be used
- Independent measures
- Scores will give data on the DV
- Matched pairs were used to eliminate intelligence as a participant variable

Apparatus

eye test and AQ test

Group 1

(Participants)

15 adult males with AS/HFA volunteer sampled with advertising through magazine and support groups

Group 2

(Participants)

122 normal adults opportunity sampled from educational classes and public libraries around Cambridge and Exeter

Group 3

(Participants)

53 male, 50 female randomly sampled from Cambridge University; assumed to be highly intelligent, with 71 majoring in science

Group 4

(Participants)

14 adults with matched IQs to Group, randomly selected from general population

Control #1

(Picture)

All pictures will be in black and white and only of the eye region of the face, taken from magazines

Control #2

(Picture)

Each card will have 4 words placed in the corners of the card, with each word based on how the person is thinking/feeling as it relates to their eye region

Control #3

(Validity)

Words on the card will be selected by two of the authors to eliminate bias

Control #4

(Validity)

Eight judges will review all 40 cards, 5 out of the 8 judges must agree to use the word choices in order for them to stay on the card

Control #5

(Validity)

No more than 2 out of 8 judges could pick the same wrong word on the card for the card to stay

Control #6

(Validity)

50% or more of the people from Groups 2 and 3 had to select the correct word of a card or it would be thrown out

Step 1 (procedure)

All participants in all groups will complete the revised eye test in a quiet room by themselves. Group 1 will be asked to identify the gender of the person on each card

Step 2 (procedure)

Groups 1, 3 and 4 will take the AQ test, all participants are asked to go through the glossary and they will have access to it throughout the test

Finding #1

The mean scores for the eyes test were 21.9 (Group 1), 26.2 (Group 2), 28 (Group 3), and 30.9 (Group 4)

Finding #2

The mean scores of the AQ test were 34.4 (Group 1), 18.3 (Group 3) and 18.9 (Group 4)

Finding #3

Females in both Groups 2 and 3 scored insignificantly higher on the eyes test

Finding #4

Males in Group 3 scored insignificantly higher on the AQ test

Finding #5

There was a -0.53 correlation between the eyes and AQ tests

Finding #6

The lowest score on the eyes test was 17, the highest was 35, and the mode was 25

Conclusion #1

All aims and hypotheses were supported

Conclusion #2

Since the data fits into a normal curve, the modifications of the 2001 test improved upon the 1997 test

Conclusion #3

The revised test was able to replicate the findings of the 1997 test and was able to identify significant impairment in the AS/HFA group

Conclusion #4

This test is now validated in identifying adults with AS/HFA

Ethics

Informed consent was obtained without use of deception. Confidentiality was maintained. Baron-Cohen doesn’t state that protection from harm was violated, it can be argued that participants may have been stressed by their scores

Strength #1

The procedure is standardized, allowing for easy replication

Strength #2

The experiment being in a laboratory makes it highly controlled

Strength #3

The use of matched pairs controls for intelligence as a participant variable

Strength #4

The use of the AQ test is a reliable and valid measure against the eyes test

Strength #5

The glossary allowed minimization of blind guessing and allowed participants to comprehend word choices

Strength #6

The numerous validity checks on the eyes test gave it high construct validity

Weakness #1

The pictures are in black and white, non-diverse, static, and only of the eye region, making the task lack mundane realism

Weakness #2

The experiment was in a laboratory, lowering ecological validity

Weakness #3

Only a simple majority was needed to approve word choices for the eyes test

Weakness #4

The correlation between the tests does not prove causation

Weakness #5

Despite the use of the glossary, participants might not have understood words

Weakness #6

There is a still a 25% of participants guessing correctly

Application

(issues & debates)

* to creating therapies for improving social communication and social emotional skills
* to schools for providing extra support in understanding emotions and reading faces