Baron-Cohen Eyes Test

Question 1

AIM 1

Answer 1

1) To test adults with high-functioning autism (HFA/Asperger syndrome-AS) on the “revised eyes test” to see if the deficits on the original test were still seen.

Question 2

AIM 2

Answer 2

To see if there is a negative correlation between autism spectrum quotient (AQ) and eyes test scores.

Question 3

AIM 3

Answer 3

To see if females scored higher on the eyes test than males.

Question 4

Research Methodology

Answer 4

Baron Cohen-et al used experiments and correlations: the scores if different groups/genders of pps were compared and the researcher also wanted to see if there was a correlation between the AQ and eyes test scores.

Question 5

Design and Variables

Answer 5

2 main variables for the experiments: the group a pp was in and sex: male versus female. Both were independent measure designs. The dependent variables were the eyes tests and AQ scores.

The co-variables for the main correlation were the AQ score and eyes test score.

Question 6

Correlation is always association, never…

Answer 6

Causation

Question 7

Hypotheses

Answer 7

AS/HFA will score significantly lower on eyes test than the control group groups.

AS/HFA will score significantly higher on the AQ than other groups.

“Normal” females, in group 2,3 will score higher on males eyes test.

“Normal” males (group 3). will score higher than females on AQ test.

Scores on the AQ and eyes test will be negatively correlated.

Question 8

The procedure was:

Answer 8

highly standardized and all pps were exposed to the same photographs, chosen using pilot studies.

It was a naturalistic/quasi experiment and pps were assigned to a condition based on a given characteristic (autism or not.

its almost a lab one-as it was performed individually in a quiet room, not a lab or natural environment)-quasi

Question 9

SAMPLE-Group 1

Answer 9

HFA/AS Adults (15males with HFA/AS):
-mean age=29.7
-mean IQ=115

Volunteer sample from adverts in UK National Autistic society magazine or support groups. All diagnosed using established criteria. Socioeconomic status and education level matched GROUP 2.

Question 10

SAMPLE-Group 2

Answer 10

General Population Controls (122 neurotypical adults)
-mean age=46.5

From adult community and education classes in Exeter or public libraries in Cambridge (Both UK), Broad range of occupation and education levels.

Question 11

SAMPLE-Group 3

Answer 11

Students(103(55males and 50females) neurotypical and undergraduate students from Cambridge University.
-mean age=20.8

Assumed to have high IQ due to very high university entry requirements,

Question 12

SAMPLE-Group 4

Answer 12

IQ-matched controls (14 randomly selected people from the general population)
-mean age=28.0
-mean IQ=116

Matched on IQ and age to GROUP 1

Question 13

PROCEDURE A

Answer 13

Pps in all 4 groups took the revised eyes test in a quiet room, alone, in either Exeter or Cambridge.
The test required them to select which of the 4 words matched the expression for each set of eyes.

Question 14

PROCEDURE B

Answer 14

Pps with HFA/AS (Group 1) were also asked to determine the gender of the eyes as a control task. The other groups did not need to do so as neurotypical adults scores have a ceiling effect. Pps were also asked if they were unsure of any words in the glossary and to read the meaning of these.

Question 15

EYES TEST Development

Answer 15

Two adults created the target words and foils for each item. These were then presented to groups of 8 judges (4 males and 4 females),who had to agree on target word for an item to pass and no more than 2 judges could choose the same foil (5 had to agree).

Question 16

A bit on the eyes test data

Answer 16

Furthermore, data for groups 2,3 were combined. For the 255 responses, at least 50% needed to select some foil. 4/40 items failed to meet these criterias, leaving 36 items on revised test.

Black white photos of diff make and female eye regions,taken from a magazine.

Question 17

Problems and solutions-
Problems with the original test and solutions in the revised new test.

Answer 17

-There were only 2 possible responses for each questions. A person would need to score more than 17/25 to ensure their score was not due to chance.
-Parents of children with autism scored similarly to people with HFA or AS, even though they did not have the conditions themselves.
-There weren’t many possible scores above chance level, leading to a ceiling effect. This meant that the test wasn’t very sensitive to individual differences.

To remedy these issues, Baron-Cohen et als new test increased number of items on test(from 25-36) and increased number of response options for each question (2-4).

Question 18

Some items were too easy as they were basic mental states (“happy” “sad”“angry”) which can even be determined by very young children: solution:

Answer 18

The new version contained only complex mental states, requiring the attribution of a belief to a person.

Question 19

Some items could be guessed from eye direction, such as “noticing”

Answer 19

These were omitted on new test.

Question 20

An unequal number of male and female faces (more females)-which may have caused bias.

Answer 20

Equal number of male and female faces used.

Question 21

The target word and the foil(incorrect answer) were always semantic opposites (e.g sad vs happy) which was too easy.

Answer 21

In the new version, the 3 foils were similar to target word to increase difficulty.

Question 22

It was unclear whether some people could understand all words.

Answer 22

A glossary included in new version.

Question 23

RESULTS 1

Answer 23

-group 1 performed significantly worse than other groups.
-Females scored higher than males, although this result was not found to be significant.
-Group 1 scored significantly worse than 3 and 4 on the AQ. Note: Group 2 did not take AQ test).
-males scored higher on AQ Than females.

Question 24

RESULTS 2

Answer 24

-No correlation between eyes score test and IQ.
-Negative correlation between AQ score and eyes test score (-0.53) for all 3 groups.
-In student group, eyes test score was negatively correlated with the social skills category (-0.27) and communication category (-0.25) of the AQ.
-Adults with HFA/AS scored 35/36 or above on the gender recognition test.