2.2 - Baron-Cohen et al.: Eyes Test
Key concepts in the study
Baron-Cohen et al.'s research focuses on understanding how people interpret mental states, particularly in individuals with autism spectrum disorders. This involves two main psychological concepts that are central to social interactions.
Theory of mind
Theory of mind refers to the ability to recognise and understand the mental states of oneself and others, such as thoughts, feelings, and intentions. This skill helps in predicting behaviour and navigating social situations.
It develops in two stages:
- Identifying a person's mental state - For example, recognising that someone feels sad.
- Understanding the context behind that mental state - For example, linking the sadness to a specific reason, like losing a favourite item.
People with autism spectrum disorders often experience a theory of mind deficit. This means their ability to fully develop or apply theory of mind is impaired, leading to difficulties in understanding others' perspectives.
Social sensitivity
Social sensitivity is the capacity to detect and interpret social cues, and to grasp the context of interactions with others. It involves accurately perceiving other people's emotions. This skill is closely linked to theory of mind, as both contribute to effective social functioning.
Background to the research
Developing accurate tests for cognitive dysfunction can be challenging, especially for adults with typical intelligence but reduced social understanding. Many existing tests are designed for children, making it hard to assess subtle deficits in adults.
The original eyes test
The original 'Reading the Mind in the Eyes' test was created to measure adult social sensitivity and theory of mind. It used 25 photographs of eyes from famous actors, where participants chose one of two words to describe the emotion shown. Results showed that adults with high-functioning autism (HFA) or Asperger syndrome (AS) scored much lower than neurotypical controls.
Problems with the original test and solutions in the revised version
The original test had several limitations, which were addressed in the revised version to make it more effective.
Here is a comparison:
| Problems with the original eyes test | Solutions in the revised eyes test |
|---|---|
| Only two possible answers per item, requiring a score of 68% or more to exceed chance. | Increased items from 25 to 36 and options from two to four. |
| Used basic mental states (e.g., happy, angry) that even young children can identify. | Focused only on complex mental states (e.g., contempt). |
| Eye direction could cue some answers (e.g., for noticing). | Excluded items where eye direction provided clues. |
| More female than male faces. | Balanced equal numbers of male and female faces. |
| Correct and incorrect options were opposites (e.g., sad versus happy), making it too easy. | Made target words and foils (incorrect options) more similar in meaning. |
| Participants might not understand some words. | Included a glossary for word definitions. |
These changes aimed to create a more sensitive tool for detecting theory of mind deficits in adults.
Aims of the study
Baron-Cohen et al. sought to refine and test the revised eyes test while exploring its relationship to autistic traits.
The specific aims were:
- To examine if scores on the Autism Quotient (AQ), a questionnaire measuring autistic traits, negatively correlate with scores on the revised eyes test (meaning higher autistic traits link to lower test performance).
- To trial the revised eyes test with different groups, as detailed below:
| Participants | Specific aims |
|---|---|
| Adults with high-functioning autism (HFA) or Asperger syndrome (AS) | Investigate if they score lower on the eyes test than other groups, consistent with the original test; and if they score higher on the AQ than other groups. |
| Neurotypical adults | Check if females score higher than males on the revised eyes test; and if males score higher than females on the AQ. |
Methodology of the study
This research used an experimental approach with an independent measures design, where different groups are compared without participants repeating conditions. It also included correlational analysis to explore relationships between variables. Data were collected quantitatively through standardised tests.
Variables
- Independent variables - These were participant characteristics: whether individuals had HFA/AS or not, and whether they were male or female.
- Dependent variable/co-variable - Scores on the revised eyes test and the AQ, which were measured and correlated.
Data collection techniques
- Questionnaire - The AQ consists of 50 closed questions answered on a four-point rating scale, assessing traits associated with autism.
- Revised eyes test - This involves 36 black-and-white photographs of eye regions, each with four word options to choose the best description of the mental state shown.
Participants
| Group | Size | Demographics | Sampling technique |
|---|---|---|---|
| Group 1: HFA/AS adults | 15 | Male adults with HFA/AS from the United Kingdom; similar socioeconomic class and education to Group 2. | Volunteer sample via adverts in the National Autistic Society magazine and support groups. |
| Group 2: General population controls | 122 | Neurotypical adults from the United Kingdom; varied occupations, socioeconomic classes, and education. | Opportunity sample from community classes and libraries. |
| Group 3: Students | 103 (53 males, 50 females) | Neurotypical undergraduate students from a selective university in the United Kingdom; high IQ. | Opportunity sample from Cambridge University. |
| Group 4: IQ-matched controls | 14 | General population matched on IQ to Group 1. | Matched sample selected to have similar IQ to Group 1. |
Procedure of the study
The study involved developing and administering the revised eyes test under controlled conditions.
Developing the eyes test
- An initial draft was tested with eight independent judges.
- Items were kept if at least five judges selected the target word and no more than two chose the same foil.
- The test was then piloted on Groups 2 and 3 (225 participants total).
- Items were accepted if 50% chose the target word and no more than 25% picked the same foil; four items were rejected, resulting in 36 final items.
Administering the tests
- Participants completed the revised eyes test individually in a quiet room.
- Groups 1, 3, and 4 also took the AQ.
- Group 1 additionally identified the gender in the photographs to check basic recognition skills.
- A glossary was available for word clarification.
Controls and ethical considerations
- Controls - All participants used the same standardised tests under identical quiet conditions, with uniform photo sizes, black-and-white format, and four options per item.
- Ethical issues - Potential psychological harm was considered, such as distress from not understanding emotions or words. Extreme scores might require referrals for further assessment, especially for those with HFA/AS.
Results of the study
The findings confirmed differences in performance across groups and supported the test's sensitivity.
Key findings:
- There was a negative correlation (-0.53) between AQ scores and revised eyes test scores across Groups 1, 3, and 4, meaning higher autistic traits linked to lower theory of mind performance.
- Group 1 (HFA/AS) scored significantly lower on the eyes test than other groups but showed no issues with gender recognition.
- Group 1 scored significantly higher on the AQ than Groups 3 and 4.
- In Group 3 (students), females scored significantly higher than males on the eyes test.
- Males scored higher than females on the AQ.
Conclusions from the study
The revised eyes test is more sensitive than the original for measuring social intelligence and theory of mind in adults. Its negative correlation with AQ scores suggests it can assess the severity of autistic traits, helping to identify subtle social deficits.
Evaluation of the study
Strengths
- Reliability - The standardised procedure, with identical photos and conditions for all participants, allows the study to be replicated to check consistency.
- Objectivity - Data were straightforward, with answers on the eyes test being clearly right or wrong, reducing subjectivity.
Weaknesses
- Validity - The static eye photographs do not capture real-life dynamics, where emotions involve movement, changing expressions, and additional cues like body language.
- Generalisability - The volunteer sample for Group 1 may not represent all adults with HFA/AS, as it only included those engaged with support groups or magazines.
Issues and debates in the research
Individual versus situational explanations
- Individual explanation - Autism represents an inherent difference in individuals, impairing their ability to interpret mental states from facial cues regardless of context.
- Situational explanation - Performance may vary by situation; for example, people with HFA/AS might better recognise mental states in quiet, low-distraction environments or with more time to process.
Applications to everyday life
The revised eyes test has practical uses beyond research. It could assess social sensitivity deficits in people with brain damage, leading to targeted support for improving emotion recognition. In schools, it might identify children needing interventions to enhance theory of mind skills, promoting better social development.