Baron-Cohen (eyes test)

Background

1985 Sally-Anne Test for autistic children. No test for adults

1997 "Eyes Task"

Theory of mind: allows people to lie, empathize, and guess the thoughts and motivations of others

Eyes Task succeeded in differentiating adults with AS and HFA from controls but suffered from psychometric problems

Psychometric Problems+Fixes

1. There were only a choice of 2 words for each set of eyes. The answer could be a 50/50 guess.

*4 words wee added in the revised test

2. Parents of children with AS/HFA scored at the same level. The test did not differentiate widely enough as scores covered a very narrow band.

*had 40 (reduced to 36) not 25

3. Ceiling effects (too many at the top end of the score range) were observed with too many people scoring too highly

*Having 36 items and 4 words aimed to remove the ceiling effect

4. Some tests were too easy

*fewer easy items

5. Some items were checking gaze direction (noticing)

*they were excluded

6. Had more female than males

*equal numebrs

7. The semantic word and foil were semantic opposites, such as sympathetic and unsympathetic. Could make it too easy

*the 3 foils were not semantic opposites, instead have a similar emotional valance. this increases the level of difficulty and ceiling effects should be removed.

8. The words may not have been understood by all participants

*added a glossary

Aim

1. To test a group of autistic adults to see if the revised version works

2. To see if there is an inverse correlation between the eyes test and the AQ test for sample of normal adults

3. To see if females have superiority on the eyes test

Predictions

1. AS/HFA will score lower on eyes test

2. AS/HFA will score higher on AQ test

3. "Normal" females will score higher than males on eyes test

4. "Normal" males will score higher than females on AQ test

5. Scores on the AQ and eyes test will be inversely correlated

Eyes Test Design

36 black and white photos of male and female eyes all the same size

4 words to describe the mental state of the person

one answer is correct

the number of correct answers is added and given an overall score

Question control:

2 authors, 5 out of 8 judges must agree on the answer

AQ Test Design

50 statements, choose definitely agree-disagree, no mid-choice participants

AQ test both reliable and valid

Method

Natural experiment/ questionnaire

Participants and Sampling Technique

Group 1: 15 male adults with AS/HFA recruited from an autistic society magazine and advertisement IQ of 115

Group 2: 122 normal adults (control group). Community classes or public libraries in Cambridge and Exeter. Mixed occupation and education levels. Only took Eyes test.

Group 3: Normal adults (Cambridge undergrads) 103 (53 males and 50 females) Much higher than average IQ

Group 4: IQ matched controls: 14 normal adults IQ matched with group 1 mean of 116. Randomly selected.

Apparatus

The AQ

They eyes test

Quiet room in Cambridge or Exeter

Design

Groups 1 and 4 were matched (IQs) but each group was independent

Procedure

1. All four groups were given the eyes test to complete in a quiet room

2. Participants in groups 1 3 and 4 were given the AQ test

Results

1. AS/HFA < normal on eyes test

2. AS/HFA > normals on AQ test

3. Woman > Men on eyes test

4. AQ and Eyes test had an inverse correlation

Strengths

Experimental validity

-measure an autistic trait not a normal one

-question criteria

-high control over extraneous variables

Reliable, replicable -pencil and paper test

Construct Validity- Quantitative data captures the entirety of the phenomenon

Weaknesses

Low ecological validity

-static pictures

-strange task