Baron Cohen - Eye tests

AIM

• TO investigate whether there are a negative correlation betwee the eye test and the AQ test scores

• to investigate if normal females (grooup 2&3) will score higher than normal males

• to investigate if the revised eye tests would work on the a group of autistic adults

Research method and design

• closed questionaires

• quasi-experiment

• independent and matched pairs design

IV

• whether had AS or HFA or not → naturally occuring → not manipulatd by the researchers

• Gender

• student

• matched IQ control group

The IV had 4 levels:

• adults with AS or HFA

• neurotypical adults from the general population

• undergraduate students

• matched IQ control group

DV

• opertaionalisation of ToM: scores on eye tests (out of 26) and AQ tests

• gender identification for group 1

SAMPLE

Group 1:

• 15 AS or HFA adults (diagnosed using established criteria (DSM/ICD))

• mean age = 29.7

• average IQ = 115

• all male

• same education level/socioeconomic class

• volunteer via advertisement/National Autistic Society

Group 2:

• 122 normal adults

• mean age of 46.5

• opp/volunteer in Exeter, public library

Group 3:

• 103 students at uni of Cambridge

• 53 females and 50 males

• mean age = 20.8

• assume to have high IQ→ control group of high IQ

Group 4: IQ matched group

• 14 adults

• (random sample) randomly selected from the general population

• average IQ = 116

• mean age = 28

• matched pair group design

(1) Procedure: Development of revised eyes test

select words and foils

• piloted on a group of 8 judges

• at least 5/8 judges had to agree that that the target words was the most suitable

• less than 2 judges could pick the foil

• items that failed had to regenerate and repilot.

(2) Procedure: Administration of the test

• test was administered individually in a quiet room at either Cambridge or Exeter uni

• No limit of time for completing the test

• practise test before the main test

• ps was given a glossary of terms and could ask for clarification

(3) Procedure: Revised eye tests

• 36 set of eyes

• black and white photos of the eye regions

• 4 words per eye photo: 3 foils and 1 target word

• target = best describe of what person in the photo was thinking or feeling

• foils = similar in valance or intensity but diff meaning

(4) Procedure: Autism Spectrum Quotient (AQ) test

• os in group 1,3,4 completed the AQ test

• self-report questionaires designed to measure the degree to which an individual possesses autistic traits

• 50 statement that ps rate on a binary scale (disagree or agree)

(5) Procedure: gender identification

• identify the gender of the person in the photograph

=> if they cannot do this → vision problem → not theory of mind

Result: aim: to investigate whether there is a negative correlation between revised eye test and AQ scores

• significant negative correlation (-0.53) between eye test and AQ across all groups

• higher AQ test scores → lower eye test scores and vice versa

Result: aim: to investigate whether normal females adults get higher score in revised eye test than normal male adults (group2,3)

• females scored higher on revised eye test in normal group 2,3

• for example in student group: female’s mean scores: 28.6 and male’s mean scores: 27.3

Result: aim: to investigate if the revised eye test could work on group of AS or HFA adults to differentiate them from the general population

• group 1 scored significantly less than other groups

• group 1 mean scores: 21.9/36 while other groups had mean scores ranging from 26.2 - 30.97

Conclusion

(1) Strength: validity

• improved validity

• improve from the previous eye test:

+ only two options for answer → to easy → revised eye test made it 4

+ revised used only complex mental states