Design approaches derived from the normal curve

28 April 2025

Some months ago, I proposed a model to explore emerging accessibility needs. As a quick recap, I observed that human abilities vary in type, and for each type, in strength. The distribution of each individual’s strength for that characteristic is likely to approximate a normal distribution. Using the mathematical approach of standard deviations, the distribution can be grouped into bands separated by each integer standard deviation value. From this we have five convenient bands of user strength for each characteristic, as shown in Figure 1.

Figure 1: Normal curve grouped by standard deviation

Each band represents a segment of the population with a particular strength of a specific characteristic, with respect to the entire population. The five bands are labeled:

“Minimal” - below -2 standard deviations
“Low” - between -2 and -1 standard deviations
“Middle” - between -1 and +1 standard deviations
“High” - between +1 and +2 standard deviations
“Extreme” - above +2 standard deviations

The “Middle” band contains the largest proportion of the population, about 68 percent. Accordingly, designing to be usable to that part of the population receives a considerable amount of attention. Designing for that middle range is also somewhat easier because each person’s strengths are not so far from the centre.

Outside the “middle” band, the “minimal” and “low” bands represent what I think we have traditionally thought of as a definition for disability. In the article above, I suggest that the “high” and “extreme” bands should be considered to be part of the accessibility space as well, as hyper-ability in some contexts can be as disabling as hypo-ability.

This model of human abilities has various applications that can help us explore emerging accessibility needs as well as better understand currently known ones. One such application is work I have been involved with over the past year with Dr. Rachael Bradley Montgomery and Dr. J. Bern Jordan on a project called the Digital Accessibility Framework. As we were considering how to build our model, the five bands defined above seemed useful, but those labels were not informative for our purposes. After much consideration and a lot of generalization, we came up with the following five types of support inspired by the bands above:

“Alternative” represents people who have minimal strength of a particular characteristic, and it is not really feasible to make material directly accessible, and instead needs an alternate solution. For example, text captions are an alternative for auditory content.
“Assistive technology” represents people who have lower than average strength of a characteristic, which can be supported by the use of assistive technology on existing content. For example, screen magnifiers help people with low vision perceive visual content.
“Default” is the way content is designed for the mainstream population without customization or assistive technology.
“Adjust” addresses higher strength or sensitivity in a particular area, where the user may need adjustments to accommodate that. For instance, a user may need to turn audio volume down (or up).
“Avoid” addresses people who are extremely sensitive to a stimulus and need to avoid it. For instance, people who have photosensitive epilepsy need to avoid exposure to triggering content.

These are extreme generalizations and in no way do we think that this is descriptive of all cases. To begin with, assistive technology users often also need alternative content. Many users across the spectrum benefit from the ability to adjust, and there are various reasons that it is best to be able to avoid certain types of content. So while these five approaches are inspired by the five bands of the normal distribution, ultimately they do not mean the same thing.

Creating these abstract categories gives us a framework to consider how to meet a given accessibility need. Is it necessary to provide alternative content, or can an assistive technology support interaction? Instead of needing to avoid something, can something be adjusted? As we look through different types of access needs, these questions help us to come up with a more comprehensive framework for accessibility in the future.

In another article I describe the Digital Accessibility Framework and describe how this concept fits with others we developed for that project. The Digital Accessibility Framework has promise to help advance the field of digital accessibility. Check that out and see how you can get involved!