Those on the autistic spectrum, how do they process information?

Researchers who studied how we process information have found that dyslexics have difficulty learning from stimuli to which they were previously exposed, while those on the autistic spectrum use prior information but update it slowly

autistic spectrum study

None of us sees the world as it really is.

When we see, hear, or smell what is around us, we do not passively absorb the information that exists in the world: the moment it is absorbed by our senses, we process the information in part based on what we have seen,heard and smelled in the past. 

For example, when our dog runs to the end of the street, we realize that it will not get smaller just because it's getting less space on our retina now. 

Similarly, we do not think the red swing in the garden changed color on a cloudy day, although objectively, the wavelength it returns is different from the one it brought back when it was sunny. 

Different perceptual biases are built into the way we process information, allowing us to use a wider context to understand raw messages that reach our eyes and ears. 

They can be examined, among other things, by optical or auditory illusions that reveal the hidden assumptions behind them.

In a new study, researchers at the Hebrew University of Jerusalem examined how perceptual biases work in people from two very different groups: people on the autistic spectrum and people with dyslexia. 

The researchers have been working on dyslexia for many years, and have developed computational methods that demonstrate how people learn environmental stimuli with simple tasks, and they were curious to see what is happening in autism. 

While subjects with dyslexia were heavily influenced by the latest stimuli exposed to them, but found it difficult to construct an orderly model of what happened earlier, subjects on the autistic spectrum were very good at constructing such a model, but slow in updating it according to recent events.

Back to the average

The experiment itself was very simple: two sounds were played to the subjects one by one, and they had to say which one was higher. Each participant listened to dozens of pairs of sounds, one after the other. 

The perceptual bias expressed here is a bias toward the center: we tend to perceive stimuli, sounds, or visions as closer to the average of a similar stimuli we saw in the past. If we see a red object, for example, we will remember it later as having a color similar to the red we saw in the past. 

In addition, the bias of sequential dependence leads us to attribute greater significance to the last thing we saw or heard.

In the experiment, these perceptual biases required the subject to relate not only to the sound it hears now, but also to those he or she have heard before, with an emphasis on those sounds heard it in the past few seconds, and to evaluate whether the frequency of the new sound is close to the average; a low note.

Previous studies have shown that this bias is stronger with respect to the first sound of the two sounds in a pair, probably because the subject should not only estimate its frequency but also remember it so that it can be compared to the second sound. 

This "additional work" burdens the system and adds noise, which is expressed by a greater reliance on perceptual bias.

When the two sounds are above the average in terms of their pitch, but the first sound is lower - the bias makes it sound closer to the average, which is even lower, so the subject will notice the difference between the sounds. 

The same is true when both sounds are lower than average, but the first sound is higher. 

On the other hand, if the first sound is higher than average, the bias makes it sound closer to the average and therefore closer to the second sound, and it will be harder to distinguish between the sounds. 

Studies have shown that differences in the ability to distinguish between these two conditions can be very large: subjects who could not determine which sound is higher than a simple guess when the first sound was far off average achieved about 90 percent correct answers when the first sound was closer to the average than the other.

Fast forgetting, slow update

When the same experimental group examined people on the autistic spectrum and people diagnosed with dyslexia, they found that subjects from both groups did not weigh the previous sounds they heard in the same way as people who were not diagnosed with these syndromes. 

The subjects on the autistic spectrum exhibited less of the sequential dependence bias, meaning they attributed less significance to the sounds they had heard in the last few seconds, but did take into account the sounds of the whole experiment just like the control group, which included non-autistic or dyslexic subjects. 

On the other hand, those with dyslexia had no problem weighing the last sounds they heard, but they did not take into account the total sounds in the same way as the subjects in the control group.

Are the difficulties shown by the subjects in the sound recognition task related to their difficulties in the real world?

The researchers believe the answer is positive. Let's begin with dyslexia. People who suffer from this syndrome have difficulty reading fast, usually due to a lack of understanding of the rules and arrangements of the written language, which are built in our minds after repeated exposure. 

So if there is difficulty to take into account the many written words a person have read in the past, and to derive these rules from them, it will be difficult for him to read quickly and efficiently. 

The researchers call this difficulty "quick forgetfulness" of what we have seen in the past, and in their opinion it points to a fundamental characteristic of dyslexia.

What about those on the autistic spectrum?

Their problem is the opposite: they assimilate the statistics of what they were exposed to in the past, but find it difficult to update the statistics, so the last things they heard or saw do not get much weight. 

Their relative disregard for up-to-date information means that they find it difficult to adapt to changing situations. This phenomenon is prominent in social interactions, in which we must act.

What are the implications of these findings? Will the findings from the study help to find ways to assist dyslectic people or people on the autistic spectrum?

The researchers believe so, but it will require adjustment to their specific abilities, rather than an attempt to change or "correct" those capabilities - which is not at all clear whether it is possible. 

According to one of the researchers, each of these populations has strengths and weaknesses. 

And therefore, it is certainly possible to adapt work or learning environments to them, which is a more promising direction than trying to teach them to do a quick update. 

When teaching autistic people, it is possible to make sure that between each new item they will have enough time to update their internal model, and so perhaps they will learn better.


  1. Lieder, I., Vincent, A. Frenkel., O., Jaffe-Dax, S., Maneesh, S., & Ahissar, M. (2019). Perceptual bias reveals slow-updating in autism and fast-forgetting in dyslexia. Nature Neuroscience.
  2. Raviv, O., Ahissar, M., & Loewenstein, Y. (2012). How recent history affects perception: the normative approach and its heuristic approximation. PLoS computational biology, 8(10), e1002731.