Cognitive Approach (Baron Cohen et al.   2001)



Psychology 9990


Cognitive Approach (Baron Cohen et al. 2001)

Cognitive Approach (Baron Cohen et al. 2001)

Background Of The Research: 

This study targets the clinically diagnosed demographic of people on the Autism Spectrum. Individuals with autism usually have hindered social functioning, communication, coping with the changes of life and also have a very specific, limited set of interests. 

“Autism” is generally used as a wide term to identify and classify a range of disorders with HFA and AS to even acute forms of autism. 

In children specifically, it is characterized by a triad of impairments: 

Communication: Verbal and Non-verbal and interpersonal.
Social interaction 
A lack of imaginative play. 

The Psychology Being Investigated: 

An important aspect that this study hypothesizes and analyzes is the “theory of mind”. What is meant by “theory of mind” is the ability to reflect and analyze one’s own and other people’s full range of mental states such as desires, beliefs, emotions, intentions and imaginations; which propagate action. It also prompts how we use this knowledge to explain and predict the actions of other people. Which by theory, is the ability to compartmentalize, understand and accept how people who are not us, are different from us, can hold opinions and emotions which differ from ours.

However, when looking at this from a psychiatric and psychological perspective, lacking this attribute is one of the key features of the Autism Spectrum, which is found generally in people with High Functioning Autism (HFA), as they find it difficult to understand that other people have their own thoughts, plans and perspectives/views. 

Baron Cohen proposes that people with autism lack or have an under-developed cognitive process/growth defined as a “theory of mind”. This theory is associated with empathy, or in this case a lack thereof (in the autism spectrum conditions). 

Baron Cohen et al. Devised a test that examines the first stage of the theory of mind known as the “Reading the Mind in the Eyes” that tests adults to use their ability to assign emotional states and labels to other people. According to him, it is a good measure of social cognition.

The two standardized and credible journals of psychiatry are the Diagnostic Statistical Manual and the International Classification of Disorders. 

In the 5th revision of the DSM, individuals who are to be diagnosed with autism have to fit into 2 criteria - patterns of restricted, reiterating behavior, activities and interests along with an inability and disparity in social interaction and communication. While the revision was in process, it was concluded that there wasn’t enough objective evidence to distinguish between AS and HFA Spectrum disorders and thus it was categorized into the “Autism Spectrum Disorder”. 


1. This study aimed to test whether a batch of adults diagnosed with Asperger Syndrome or High-Functioning Autism  (HFA) would be impaired on the revised version of the “Reading the Mind in the Eyes” test/task. 
2. To investigate and test if there was a relationship between a participants’ performance on the revised “Reading the eyes in the mind” test and degree of autistic characteristics. 
3. To investigate if there were sex differences in those without autism on the revised “Reading the mind in the eyes task”. 


1. Participants with autism significantly lower scold on the revised reading the mind in the eyes test in the control group. 
2. Participants with autism will score higher (significantly) on the Autism Spectrum Quotient Test (AQ) measure. 
3. Females in the ‘normal’ groups will score higher on the “Reading the Mind in the eyes test” than males in those categories.
4. The “Normal group” which consisted of males would score higher on the AQ measure than the females. 
5. The scores of the ‘Reading the Mind in the eyes’ test and the AQ measure would be negatively correlated. 


Research Method: 

- It was conducted in a Laboratory as the situation put forth in the experiment lacked mundane realism (as they were asked to identify and categorize emotional information from photographs).
- This was a highly standardized procedure, as all the participants in the study were shown the same set of pictures. (Which were chosen from a pilot study). 
- The study was however identified as a quasi-experiment (where the variable is naturally occurring).
- The Independent Variable: The independent for the room was the kind/type of participant in each given condition. Three comparison or control groups comprised the study along with the experimental group which contained the participants with AS or HFA.
- The Dependent Variable: 
- 1. Score of the revised “Reading the mind in the eyes task”. The participants in the AS/HFA condition and IQ matched control condition, there was also an AQ test. 
- 2. Scores of IQ tests of AS/HFA participants vs IQ scores of the matched-IQ group. 

- Utilizing the “Reading the mind in the eyes” test in 1997, Baron Cohen et al. conducted an experiment to investigate whether adults with HSA or AS had problems using a theory of mind.
- To test this Baron Cohen et al. compared a group of individuals who had HSA or AS with some who didn't and the individuals were shown various pictures of eyes and asked to identify the emotion that was being emitted from those two options. This is what was known as the Reading the mind in the Eyes task. 
- In order for the individuals to better describe the emotions shown in the pictures, for there to be reliability and standardization, and for there to be validity in the details shown in the pictures there was also a glossary used. 

- Developing the test: 
- The test consisted of target words and foil words (which were opposites) which were selected by the first two primary authors and were then piloted on 3 categories of 8 judges, with an equality of gender- 4 males and 4 female judges. 
- The approval criteria decided was that at least five out of eight judges would agree that the target word was the most suitable description for each stimulus. Furthermore, only 2 judges picked a solitary foil. 
- Items that did not succeed in meeting the criteria had been assigned the task of new target words, foils or both designed and were then repiloted with following groups of judges till the criteria was met for all items.

Reading the Mind in the Eyes Test (Original)Reading the Mind in the Eyes Test (Revised)

  • This version had only 2 responses for the participants to choose from - two words presented per picture only.
  • When the parents of the participants with AS, they too scored below the general population. (As autism is also genetically interspersed.)
  • When the narrow range of the acquired responses, which were significantly above odds on the first testing can lead to a score which lied in the ‘normal range’ - being close to the ceiling of the test, lowering detection of participant variables. 
  • This version included both rudimentary and complex mental states and some items were quite basic to detect again increasing chances of closing it to ceiling effects. 
  • Some items could be solved very easily by just the direction of the gazes ( in the picture ). 
  • This version had an increased ratio of female pictures and this may have made it gynocentric- unrepresentative to the general population. 
  • In this version, pairs of the words were always semantic opposites (for instance; reserved vs unreserved and bothered vs bothered), limiting the range of response from individuals and making the test too easy. 
  • In this revision, it was uncertain if there were comprehension problems with the words, which might have affected an individual’s score, because this included mapping the words to the picture. This is specifically concerning for individuals in the HFA group, where a language delay was probable. 
  • The items on the test increased:
  • No. of pictures increased from 25 to 36. 
  • No. of response options increased from 2 to 4, to effectively indicate participant variables and less chances of ceiling effects. 
  • The rectification showed; that parents of individuals with AS also scored lower than the general population.
  • The inclusion of both rudimentary and complex mental states now added in the complex states which required to attribute intention or belief, so increase in validity. 
  • Now equal number of men and women’s faces in revised version, thus eliminating risk of gender bias - but allowing to test gender differences. 
  • Gaze direction was sufficient to solve some items (noticing/ignoring), hence this part of the test was ruled out in the revision. 
  • Because target and foil words were too simple, the foils were designed in similar emotional valence as of the target words which increased complexity. For eg; ashamed/alarmed/anxious for “serious”.
  • A glossary was included to assist the participants with linguistic insufficiency as reference, instead of having to map the words to the picture single-handedly.

Participant Groups (Sample): 

Group #1: AS/HFAGroup #2: Adult Comparison groupGroup#3: Student Comparison groupGroup#4: IQ Matched group 

- 15 male adults

- Mean IQ score: 115.

- Mean Age: 29.7 years.

- Self-selecting sample, as they were chosen from support groups/Autistic Society Mag. 

- Diagnosed by DSM or ICD.

- Normal functioning adults, with NO AS/HFA condition. 

- Mean age: 46.5 years.

- Sample selected from Public Library and classes being conducted in Exeter. 

- Normal functioning students, with NO AS/HFA condition.

- Mean age: 20.8 years.

- Selected from Cambridge University. (Non-representative sample because it is a very specific university for the larger context. 

- 14 participants with NO AS/HFA. 

- Mean age: 28.

- Mean IQ: 116. (Normal people)

- Random participant selection. 


- After the changes has been made to Original Eyes Task, the Revised task was used in this experiment as a measure of the theory of mind. 
- This included 36 pairs of eyes, each with 4 options of emotions to identify the picture with, on the face of the target. (The test was individually administered in a quiet room in Exeter or Cambridge). Ample time was given to assess the options.
- Individuals in all 4 of the groups were made to take the Revised Eyes Test and additionally, the AS/HFA participants were also asked to determine the sex of the target in each picture - this was a control condition to optimize any considerable predictions. It was anticipated that the AS/HFA patients would have mental state recognition impairments. 
- Initial testing had shown that other groups performed sufficiently on sex-recognition tasks, so they were not to take the test.
- Other groups 1,3 and 4 completed the AQ measure. (Group #2 did not)
- Every participant went through the glossary of terms to understand what each word meant. They referred back to it as reference if they wanted to, in the middle of the experiment.
- All participants had consented to be included in the study, being aware of the nature of the research. 
- AS/HFA individuals were clinically diagnosed viaAPA criteria. 
- Baron Cohen et al.  retained anonymity so no individual would be identified via their scores, hence privacy and confidentiality was maintained. 


- On the Eyes Test; the AS/HFA participants accurately identified significantly lesser target words than participants in other comparison groups. 
- In comparison groups (adult and student), findings of sex differences were evident between females and males on the Eyes Test. 
- All individuals in the AS/HFA condition scored 33 or above on 36 of the sex recognition tasks. 
- As expected, on the AQ group #1 - males scored higher than females in group 3.
- Individuals in the AS/HFA condition scored significantly higher than the student comparison and IQ matched group.
- Subjects in all 4 groups didn’t differ in the number of words which they were unsure of, in the glossary. 
- In all the subjects, the number of words checked never went higher than 2. 
- A distinctive negative correlation was found between the Eyes Test and AQ scores. 
- However, no correlation was found between IQ and Eyes Test scores.
- This is indicative of how if a participant’s AQ score increases, their ability to accurately verify the target word on the Eye Test decreases. 


- The findings of the study Propose that people with AS or HFA have a cognitive deficit which does not allow them to process information using the theory of the mind, so that they may identify emotions in other individuals.
- This lack of the theory of mind is strongly linked to autism spectrum disorders.
- Evidence suggested that there was a difference of sex between males and females in the comparison groups; with men showing more tendency of autistic traits and performing worse on the ice chest then females.
- However some of these differences were insignificant and unsubstantiated, so further research would be needed to clarify if there were any differences between sexes and their AQ scores.
- It is obvious that the revised version of the Eyes Test utilized in the study was a more acute and sensitive measure of adult social intelligence than previously used in studies. If there are future studies on adult social intelligence participant variables, may exist but in a more constructive way.
- The correlation between IQ score AQ score and eyes test scores was computed. When the groups were combined, there was no correlation between the IQ measure and Eyes Test or between the IQ or the AQ.
- There was a relationship of inverse proportion discovered between the Eyes Test and the AQ measure. This was accurate for all three groups report measures were employed.
- A relationship of inverse correlation was also found in the student group where the Eyes Test was inversely correlated with the social skills category in the communication category.

Strengths And Weaknesses:

  • The experimental setting was in a Laboratory, so the confounding variables were strictly controlled. High internal validity.
  • The study could be replicated and thus reliability could be proved.
  • Quantitative data was gathered, in numeric form in the IQ and AQ test scores - thus easier to statistically analyze and organize. 
  • Revision of the Eyes Test meant improved validity, which decreased chances of ceiling effects. 
  • Ethical considerations were kept in check.
  • Because it was conducted in a lab setting, it had low ecological validity and thus lower mundane realism. 
  • This was a natural/quasi-experiment, where it wasn’t possible to randomly assign a participant to a group, to lower possible order effects ( as the participants in a certain group confronted a certain kind of questions) and thus the very nature and validity of the study maybe questioned. 
  • No concrete explanation of the results conceived. 
  • Ethical consequences- some participants may feel stigmatized after taking the Eyes Test. This may lead to psychological problems and affect social functioning.
  • Because the sample was very specific, it is not possible for the findings to be applied to everyday life, cannot be generalized to the general population. 

© 2021 O’Level Academy. All Rights Reserved