Myths About Low Vision
Most people lump blindness and visual impairment into one group. This is a mistake that does serious harm to many people who have low vision but are not blind. Well meaning people cite accommodations for people who are blind as examples of things that work for all people who are blind or visually impaired. Even experts do this too. This includes many advocacy groups, national, regional and local governments, institutions and even the W3C WCAG Working Group. These groups recognize and address the needs of blindness and non-visual readers, but they frequently exclude the most critical needs of visual readers with low vision from standards and policies.
There are two groups of people with low vision: Visual readers with Low Vision (VR/LV) and non-visual readers with low vision (NVR/LV). Many accommodations that serve the blind also serve NVR/LV very well. Reading needs of VR/LV are very different. Accommodations for people who are blind or NVR/LV frequently do not help VR/LV and are even impediments.
Low vision accounts for more than half of the people who are blind or visually impaired. VR/LV is a huge sub-population of people with low vision. That means when you get something wrong for VR/LV, you are harming lots of people.
Whenever you think about low vision you must remember that you are not addressing one disability; you are addressing a large cluster of disabilities. Something that works for one population, has a good chance of impeding another population. Regarding print, visual readers need individualized accommodation. This is best exemplified by HTML with CSS that has fully separated information and structure from presentation. Using the unfettered power of CSS most people in VR/LV can have their reading disability substantially if not completely removed. Print disabilities for visual readers with low vision are avoidable today. They only persist because publishers and media vendors persist in using media that prevent individualized accommodation.
To address this issue we must dispel the myths that encourage standards and policy bodies to quit working after they have solved the problems for the blind and NVR/LV.
People with visual impairments need text alternatives for non text content.
False —an accessibility feature for blindness that does little for VR/LV: Alternative text is invisible to VR/LV in most browsers. To see the content of the ALT attribute you must turn off images. This will hide even the images the visual reader can see and understand.
Images are a great place to use screen magnification (Zoom). This helps with small images and tiny buttons in forms and image maps. Since the alt-text does not display visually in browsers to help visual readers with low vision, the only way to know the purpose of an icon is to zoom it.
Tool tips are very useful for VR/LV. You can often make them big, but ALT text does not show up as tool tips. Most browsers use the TITLE attribute for tool tips.
The TITLE attribute is very useful for VR/LV. Sadly, many developers suppress the TITLE attribute in a misguided belief they are helping people with visual impairments. This harms visual reading with low vision.
People with low vision use screen magnification for their primary accommodation.
False in general: As mentioned above low vision is a cluster of disabilities. Some people with low vision do not use enlargement at all.
For those who use enlargement, visual readers with low vision use screen magnification as an accommodation of last resort. If they can obtain materials in a typographic format that meets their reading needs, they pick that format first.
Without word wrapping, magnification has limited value. Navigation without word wrapping is a struggle. Screen magnification software forces this navigation mode in many cases.
Some magnification software enables linear text streams and other continuous access methods. This works well except when the content of text requires considerable cross referencing as in quantitative content.
Screen magnifiers can be great navigation tools. People with low vision use magnification for reading buttons, small icons and fixed size fonts. Screen magnifiers are some of the best navigation and operation tools available.
Unfortunately, leading screen magnifiers are very intrusive; they are rigid and hog screen space. They focus on needs of people with severe to profound visual impairment. People in VR/LV avoid this software, or keep it in the background only to be brought out for spot use. The high cost of leading screen readers is not justified for this type of use.
Standard browser enlargement, with word wrapping is widely used when possible. Style templates in word processors make reading much easier. CSS is excellent for HTML that is primarily reading content. It is less effective on more complex pages. Most users find CSS too difficult to create. Given a style sheet that is built to their needs, most people in VR/LV will use it for reading articles.
The take away here is this: Most in VR/LV find magnification is a necessary evil for reading. When you can get better content you use better content. Unavailability of appropriate content forces high use of magnification. When nothing else is available magnification sometimes works, but magnification can be to difficult to use on terse content. Math, for example, takes a lot of concentration. The extra cognitive cost of managing magnification tools can remove too much attention from understanding the content.
Large print with word wrapping fixes the problem
False almost always: The variation in this disability rarely responds to just one typographic fix.
People with loss of peripheral vision, may have excellent visual acuity. Their problem is a narrow field of vision. Large print may impede people in this group.
For those who need large print, large print with word wrapping is seldom enough. It is necessary for most of VR/LV but not sufficient. Without large print, there would be few visual readers with low vision; they just need more typographic help.
A font with serifs like Times New Roman can be visually confusing. Black print on white paper almost always causes strain and discomfort. Both high contrast and low contrast can be a barrier, depending on the type of low vision. Line, word and letter spacing can improve perception of the content in lines, the separation of words, and identification of the letters appearing in words. For people with limited visual field, line length can be a problem. Many people with low vision like boundaries separating document elements. Ignoring any one of these issues can cause discomfort and / or reading errors. This reduces stamina and comprehension.
People with visual impairments need keyboard access.
False for most visual readers with low vision: Keyboard access is necessary for blindness. NVR/LV is split on this.
VR/LV consists of mostly mouse users. Vision is a dominant sense. If you have it you use it. Reach and grab is a natural behavior for people who can see. The mouse models this activity. Not surprisingly, visual readers use the mouse.
A person with low vision who uses a screen magnifier will often enlarge the screen and find controls visually. Again, the mouse is the best tool for this.
The belief that all visual impairment needs keyboard access, makes screen reader use difficult for VR/LV. Leading screen readers focus so much attention of keyboard access that mouse use is frequently difficult. It is an after thought. This is why many if not most visual readers with low vision do not use the leading screen readers. The focus is so exclusively on the needs of blindness, people in low vision are forced to use an access mode that is not natural to them.
The inability to get help from leading screen readers is a big problem for VR/LV. Listening to long documents is a major relief for stamina, the most serious issue facing visual readers with low vision. People in VR/LV are frequently criticized or dismissed as uncooperative when they refuse to use readers that do not help them.
If you see the print in a document, then you can read the document.
False —almost universally: Stamina is the biggest issue facing visual readers with low vision. Reading errors and extreme discomfort make it difficult to finish documents. While people with blindness cannot read print at all, the reading problems of VR/LV are more subtle. One can usually start a document, but finishing is very difficult. With enlargement alone one can frequently read one or two pages. Beyond that, serious fatigue sets in.
What goes wrong? Here are a few scenarios.
Eye strain: Eyes just get tired. In time they hurt. This can build up in one sitting, or set in on the second or third day of reading a long document. At some point it hurts too much to continue.
Neck and back pain: Visual reading with low vision usually involves sitting close up with poor posture. Nose-to-the-page is often the way to read. Again you quit when the pain builds.
Nausea: If you have ever had your pupils dilated and then had the doctor shine her flood light inside your eye, you know the nausea caused by visual stress. White background with black print can produce the same nausea. It is less intense at first, but over a long time it can be just as painful and nauseating. Eventually, you quit reading.
The take home from this is: Visual reading with low vision impairs the reader over the long run. In a professional environment not finishing your reading can get you fired. Unless these readers get the typographic support they need, professional employment is out of reach for most.
Once you have fixed accessibility for reading with blindness, the job is half done.
Standards bodies and policy makers will have to rethink their approach to low vision accessibility. Most of the corner stones of accessibility for blindness help visual readers little or less than none.
This group may be more difficult to accommodate because they require very flexible data. HTML with CSS has already proven that it can be done. Web developers must take care to make sure their web content truly separates presentation from meaning. Proprietary data formats must also do the same.
Visual readers need data that they can start and finish. If they cannot get this access they will continue to be underemployed. This would be a mean and unnecessary waste of human talent.