Colour me compliant

Peter Irons, head of research at Eyescience-Tintavision, looks at the use of colour to help dyslexic adults or any person who finds reading too slow (less than 160 words per minute), painful or tiring.

Computer Vision Syndrome, as it is sometimes referred to, is rarely solved by adherence to the Regulation 5 of the Health and Safety (Display Screen Equipment) Regulations 1992. These regulations were set up at a relatively early stage in software, monitor design and function.

Despite dramatic improvements in the technology many millions of computer users are still experiencing major difficulties when using their computers.

The modern monitor, though, works because it taps into the biology and 'computing' of the visual system and it can be bespoked to the benefit of the user.

The range of colour options offered on websites concerned with dyslexia, and the range of websites offering a colour choice, indicate that most people who are dyslexic and most professionals working directly with dyslexic people in the field, accept that changing the colour is likely to be at least a good start.

Most of the people who are offered the use of colour do so from a very limited choice. A recent trend has been to suggest that there is research evidence to support the notion that the use of a yellow background assists people. Evidence supporting this comes from research with children learning to read rather than with adults.

Which colour should a person use? A colour chosen by the person; by another person from a limited choice or calculated for the person based on their measured reaction to measured colour.

This article is primarily concerned with a very precise approach to the use of colour. It makes use of the ability to:

  1. Accurately control the colour background on a computer screen.
  2. Accurately measure the effect of colour background on their reading. 
  3. Subsequently calculate the best setting for the individual.

Reading speed changes 

The graph above shows the way in which the reading speed, for a particular student, changed with precise control of the colour background. During this testing there was no change of colour seen by the student. The screen appeared as cyan at all tests.

Near to the person's optimum settings, very slight changes in the red setting, dramatically affected reading speed (this person also required the overall brightness to be carefully managed as well as the font size).

This demonstrates that the changes in reading speed are not specifically about what most people understand as colour, instead it is about the biophysics of seeing colour.

We have repeated this process of finding optimal settings with over 9,000 undergraduates at universities across the UK. We are then able, from analysis of the database, to start to see patterns and the implications for others.

Is colour the solution to problems that dyslexic adults and others have with reading?

We need to consider which specific difficulties we were working on. What are the problems? Can the problems be measured? Can we measure improvements?

Compared with the non-dyslexic students with whom they are competing, the common difficulties for dyslexic students that this work has addressed are:

  • Poor fluency in reading
  • Low reading speed
  • Poor reading stamina
  • Visual stress.

For most adults we have seen, the most emotionally debilitating and frustrating difficulty is the fluency problem, whether reading aloud or with their inner voice. They are aware of and are often told by others, that whilst reading aloud, the words do not flow. They are often embarrassed at the response of others.

For very many, prolonged reading is linked to increasing eye difficulties, excessive fatigue and then extended periods of recovery.

The important point here is that these eye difficulties were normally not opthamological, or orthoptic (muscular) weaknesses, they are not about visual health. If they did have such problems, they had been through treatments and interventions that had made no significant difference to the difficulties.

Extensive research has shown no association between functional vision and reading difficulties. In most cases they experienced the difficulties only when they were reading or writing. They did not have difficulties with diagrams and pictures.

The eye difficulties were task dependent and appear to be strategic responses to the reading task. Most did not have a visual difficulty/deficit.

It was their visual systems that could not cope with the stresses of the reading/writing process: the process of controlled, iterative visual data collection and processing. A proportion though were generally light sensitive, in that they experienced visual stress in normal vision activities in brightly lit environments.

What sort of results have been obtained?

Reading aloud on a white background

Average rate of reading (rapid automated naming, RAN) amongst the dyslexic students: 134 words per minute.

Average RAN amongst the non-dyslexic adults (academically successful): 179 words per minute.

Our data suggests a correlation between reading speed and the level of academic success. When reading on a white background, that is a differential in RAN between the two populations of 34 per cent.

Why such a strictly measured approach?

A computer programme has been developed to enable standardization and to minimize the time required to assist each person. It can be easily seen that the benefits of setting the screen up as precisely as possible has disproportionate benefits. Mathematically it is very similar to tuning in a radio, achieving resonance.

What happened using the measured approach?

By calculating the optimum settings for each student there have been greater benefits in the measured outcomes we were trying to change and this has benefited a much greater proportion of the dyslexic students.

By tracking the students over time and recalculating the optimum, we were also able to analyse the changes in individuals including how they responded to the intervention. For the first time ever, this sort of data has been recorded and analysed.

These are changes in:

  • Reading speed
  • Oral reading fluency 
  • Relative sensitivity to red, green and blue pixel brightness
  • Overall light sensitivity.

Not only were the measured benefits in reading speed and fluency much greater, they were more persistent and accumulative using the new optimal setting, and in addition there were clear benefits when reading on a white background.

Reading aloud on the first measured optimal background

Average RAN amongst the dyslexic students: 178 words per minute

For each individual there are mathematical relationships between the output variable (reading speed) and the input variables (values controlling the appearance of the background). We have here distinct independent variables, which we control, and dependent variables, which we measure. Very much like any electronic/logic based system
Also for the first time we have accumulated sufficient data on adult dyslexic people to be able to make real comparisons between them and the people who do not have any recognised dyslexic disability.

Eye movement

In addition we have data from sophisticated, very reliable, eye tracking equipment, showing the patterns of eye movements during reading on white backgrounds and optimal backgrounds.

These consistently show that the eye movement anomalies that occur when many dyslexic people are reading on a white background do not occur when using the optimal conditions.

Eyes take about 160 pictures per minute. That is they take a series of pictures of the scene being looked at. The series of pictures is organised such that each picture in the sequence is taken of a new place in the scene and at a place determined during the taking of the preceding picture.

The brain then processes the sequence of pictures into the illusion of a single scene. Confusion (sequencing problems?) arises when the eyes move to the wrong place or  move too much during the taking of a picture. The brain then finds it too hard (not enough RAM!) to computer enhance the images into a coherent scene.

With reading the sequence has to be very consistently accurate if the person is to chunk the visual data together and minimize the amount of working memory needed to achieve fluency.

With adults who find difficulties working on computer screens or with extensive reading, the eye movement is often rather disorganised compared with other adults.

When the optimal conditions are used then there is little distinction between the eye movements of the two groups. When we look at the eye movement management during graphic activities we do not usually find any anomalies.


Lack of precision can easily give rise to minimal benefits or no benefits and can leave people with difficulties, which could have been resolved. The improvement in reading fluency is substantial. The benefits build up from repeated optimization.

Peter Irons is a biologist who worked in secondary education for 28 years and since 1997 has been working with dyslexic undergraduates and adult computer users on the relationship between biology and reading difficulties.

July 2006