Baron-Cohen et al. (Eyes Test)

This study only makes sense once you understand theory of mind — the part textbooks often state in one sentence.

Theory of mind (ToM) is the ability to attribute mental states — thoughts, beliefs, intentions, emotions — to other people, and to understand that others' mental states can differ from your own. It is the basis of everyday social skills: reading a facial expression, sensing sarcasm, knowing when someone is upset.

It develops in childhood. Classic false-belief tasks (e.g. the Sally-Anne task) show most typically developing children pass by about age 4 — they understand that someone can hold a belief that is false (e.g. looking for an object where they think it is, not where it actually is).

Autism and theory of mind. Baron-Cohen's earlier work proposed that autism involves a delay or difficulty in theory of mind ('mindblindness'). But by adulthood, many people with high-functioning autism (HFA) or Asperger syndrome pass simple false-belief tasks — so a more sensitive, subtle test was needed to detect any remaining difficulty.

Problems with the original (1997) Eyes Test that the 2001 revision fixed:

• It offered only 2 forced-choice options → easy to guess; ceiling effects (everyone scored high).
• Some items could be solved by working out gaze direction, not a mental state.
• There was an imbalance of male/female faces and of basic vs complex states. The revised test used 36 items, 4 options each (1 target + 3 foils), more subtle mental states, a balanced set of faces, and a glossary of the words — making it sensitive enough to reveal subtle individual differences.

Psychology being investigated: theory of mind and social sensitivity.

Aim. Baron-Cohen et al. had several linked aims:

1. To test whether the revised Eyes Test could measure subtle individual differences in theory of mind among typical adults (and overcome the original test's ceiling effect).
2. To test whether adults with high-functioning autism (HFA) or Asperger syndrome score lower than typical adults — i.e. whether theory-of-mind difficulties persist into adulthood.
3. To check whether Eyes Test scores correlate with autistic traits measured by the Autism Spectrum Quotient (AQ) and whether there is a sex difference.

Samples (four groups):

• Group 1: 15 adults with HFA or Asperger syndrome (13 male, 2 female), recruited through a support magazine/clinics.
• Group 2: 122 general-population adults (e.g. drawn from adult community/evening classes).
• Group 3: 103 Cambridge University students (assumed to have high IQ).
• Group 4: 14 randomly selected general-population adults, IQ-matched to Group 1 (a control for IQ).

The revised Eyes Test: 36 black-and-white photographs of just the eye region of faces. For each, participants chose one of four mental-state words that best described what the person was thinking or feeling. A glossary of all the words was provided so misunderstanding a word would not lower the score (boosting validity). Maximum score = 36.

Additional measures:

• The AQ (Autism Spectrum Quotient) — a self-report of autistic traits (completed by the autism and student groups).
• A gender-recognition control task — judging the sex of the eyes — to check that the autism group were not simply worse at the general task of looking at eyes (if they did fine on this, any Eyes Test deficit is specifically about theory of mind).

• Group difference. The autism/Asperger's group scored significantly lower on the Eyes Test — a mean of about 21.9 out of 36 — compared with the control groups, which scored roughly 26–28.
• Sex difference. Among typical adults, females scored slightly higher than males on the Eyes Test (consistent with the idea of greater female social sensitivity).
• Correlation with AQ. There was a negative correlation between Eyes Test score and AQ score: the more autistic traits a person reported, the lower their Eyes Test score.
• Gender-control task. The autism group did not differ from controls on judging the sex of the eyes — so their deficit is specific to reading mental states, not to the task in general.

• The revised Eyes Test is a sensitive measure of adult theory of mind / social intelligence — it detects subtle individual differences and avoids the original test's ceiling effect.
• Adults with HFA/Asperger syndrome have theory-of-mind difficulties that persist into adulthood, even when they can pass simpler tasks.
• Theory of mind exists on a continuum, correlating with autistic traits (AQ) and showing a small sex difference in the general population.

Strengths

• Standardised psychometric test → replicable, reliable and produces quantitative data that can be compared across groups.
• Strong validity controls — the glossary (so poor vocabulary doesn't lower scores) and the gender-control task (so the deficit is shown to be theory-of-mind-specific). The revised design also removed the original's ceiling effect.
• Useful application — a tool for research into, and screening/support for, autism.

Weaknesses

• Small, unrepresentative autism sample — only 15 adults (mostly male), so findings may not generalise to the whole autism spectrum (or to women with autism).
• Low ecological validity — judging emotion from static photos of just the eyes, with no voice, body language or context, is not how we read minds in real life.
• Possible cultural bias — the faces and English mental-state words may suit some cultural groups better than others (cultural differences), and the student group may not represent the general population.

Issues & debates

• Application to everyday life: assessment and support for people with autism; understanding social difficulties.
• Individual differences: the study is built around measuring individual differences in social cognition.