Accessibility

[gledder]

I recently heard that all UNL administrative units are going to have to change their web pages to conform to the UNL template. In searching for information about the UNL template, I discovered that "accessibility" is one of the criteria for the template. This makes me wonder how "accessibility" is defined.

My daughter is legally blind, but she is able to read by holding her head close to the reading material. She is also colorblind. As a high school senior, she has been looking at a lot of college web sites. She finds the UNL web site to be one of the worst for her to use.

The fonts used on the UNL web site are very small. Yes, she can make them bigger on her computer, but that also makes everything else bigger, which changes the layout and makes the page hard to navigate. The links, which are important, are smaller than the press releases, which are relatively less important. For a visually-impaired person, it is difficult just to locate the menus, much less to use them, because the eyes are naturally drawn to the press releases that appear in a larger font.

The biggest problem is the lack of contrast in the color scheme. Some of the menu items are light gray on dark gray or dark gray on light gray. When my daughter looks at these, she cannot tell that there are any words written there. If they had been red on light grey, for example, they would have been readable. As they are, she has to have someone else control the computer when she navigates the UNL site. The UNL web site is not accessible for her.

I guess it all depends on what you want. If you want to define "accessibility" narrowly as "suitable for software that turns content into speech," then I suppose the UNL web site is accessible, although I suspect it is still difficult to find the needles of the menu in the haystack of total content. This may be satisfactory in a legal sense, but it is not satisfactory in a moral sense. If you define "accessibility" more broadly as "readable to the largest number of people using the greatest variety of tools," then the UNL template leaves a lot to be desired.

It is bad enough that the main web site has this problem. Forcing academic units to make their web sites less accessible than they currently are so that they conform to the UNL standard is wrong, and I will urge my academic unit not to comply.

[rcrisler1]

I recently heard that all UNL administrative units are going to have to change their web pages to conform to the UNL template. In searching for information about the UNL template, I discovered that "accessibility" is one of the criteria for the template. This makes me wonder how "accessibility" is defined.

My understanding is that all new web development at the UNL and Unit levels after fall 2006 will be required to comply with UNL identity policies, which will be amended at that time to require that websites use the UNL site templates. This does not include pages at the Individual level, such as your personal faculty page and others at that level. A new generation of UNL site templates will be released at that time. They are currently being developed by a subgroup of the UNL Web Developer Network representing a wide range of colleges, departments and administrative units.

My daughter is legally blind, but she is able to read by holding her head close to the reading material. She is also colorblind. As a high school senior, she has been looking at a lot of college web sites. She finds the UNL web site to be one of the worst for her to use.

The fonts used on the UNL web site are very small. Yes, she can make them bigger on her computer, but that also makes everything else bigger, which changes the layout and makes the page hard to navigate. The links, which are important, are smaller than the press releases, which are relatively less important. For a visually-impaired person, it is difficult just to locate the menus, much less to use them, because the eyes are naturally drawn to the press releases that appear in a larger font.

The biggest problem is the lack of contrast in the color scheme. Some of the menu items are light gray on dark gray or dark gray on light gray. When my daughter looks at these, she cannot tell that there are any words written there. If they had been red on light grey, for example, they would have been readable. As they are, she has to have someone else control the computer when she navigates the UNL site. The UNL web site is not accessible for her.

We will take these issues into account in the redevelopment process.

Acceptable contrast of colors, under Web Content Accessibility Guidelines, is determined by an algorithm, as follows (this algorithm is still published as a draft proposal, but it has been unchanged since April of 2000):

Color brightness is determined by the following formula:
((red value X 299) + (green value X 587) + (blue value X 114)) / 1000
Note: This algorithm is taken from a formula for converting RGB values to YIQ values. This brightness value gives a perceived brightness for a color.

Color difference is determined by the following formula:
(maximum (red value 1, red value 2) - minimum (red value 1, red value 2)) + (maximum (green value 1, green value 2) - minimum (green value 1, green value 2)) + (maximum (blue value 1, blue value 2) - minimum (blue value 1, blue value 2))
The range for color brightness difference is 125. The range for color difference is 500.

Using that W3C formula, the red over light gray that you suggest, if by light gray you mean the hexadecimal color #8A8A82 in our current palette and by red you mean the #CC0000 in our current palette, do not pass ... the color difference value using that color combination is 334, the brightness difference is 76.092. To be acceptable according to Web Content Accessibility guildelines, the former value would need to exceed 500, the latter 125. To that end, we added a high contrast option, invoked by clicking an ISO-standard contrast icon, in the templates early in 2004 that pushes all of the values to a fully compliant state. That method, allowing the user to change the color combinations, is one of the repairs suggested by the Worldwide Web Consortium. Our new generation of UNL templates, currently under development, will not require that mechanism.

I guess it all depends on what you want. If you want to define "accessibility" narrowly as "suitable for software that turns content into speech," then I suppose the UNL web site is accessible, although I suspect it is still difficult to find the needles of the menu in the haystack of total content. This may be satisfactory in a legal sense, but it is not satisfactory in a moral sense. If you define "accessibility" more broadly as "readable to the largest number of people using the greatest variety of tools," then the UNL template leaves a lot to be desired.

When we refer to accessibility, we are referring to the standards and recommendations set forth by the federal government (Section 508) and the Worldwide Web Consortium's Web Content Accessibility Guidelines. We encourage all developers at UNL to test their sites against these standards using the tools at http://validator.w3.org and http://www.contentquality.com/. The templates, as distributed, validate to the highest of these standards. UNL's was the first site among the Big 12 and our designated peer institutions to attain compliance with Section 508 of the Rehabilitation Act, as well as compliance with WCAG Priorities 1, 2 and 3. It is up to the individual developer to maintain that level of accessibility compliance in the code that is added to the templates.

It is bad enough that the main web site has this problem. Forcing academic units to make their web sites less accessible than they currently are so that they conform to the UNL standard is wrong, and I will urge my academic unit not to comply.

As noted above, we are in the midst of a comprehensive redesign of the UNL site templates. While it is too late to join the groups that have been formed to develop design prototypes, I would encourage personnel from Mathematics to be a part of that discussion through the Web Developer Network. I will note that before we began a drive toward better accessibility on the UNL website, no individual UNL college's website passed any level of HTML validation or accessibility compliance. If templates are adopted as recommended, I expect this positive trend to continue. Please contact me or the Mathematics web developer(s) if there are any further issues you'd like to see addressed in this redesign.

[gledder]

We will take these issues into account in the redevelopment process.

This gives me hope that the next version will be better.

Using that W3C formula, the red over light gray that you suggest, if by light gray you mean the hexadecimal color #8A8A82 in our current palette and by red you mean the #CC0000 in our current palette, do not pass ... the color difference value using that color combination is 334, the brightness difference is 76.092. To be acceptable according to Web Content Accessibility guildelines, the former value would need to exceed 500, the latter 125.

Does the current scheme satisfy these standards? If it does, then (at least for my daughter) the standards don't seem to be very good. In "suggesting" red over light gray, I was merely looking for an example that was consistent with the overall color scheme and which my daughter judged to be significantly better than what is there now. Maroon on white or dark blue on white would be much better, of course.

Color brightness is determined by the following formula:
((red value X 299) + (green value X 587) + (blue value X 114)) / 1000
Note: This algorithm is taken from a formula for converting RGB values to YIQ values. This brightness value gives a perceived brightness for a color.

Color difference is determined by the following formula:
(maximum (red value 1, red value 2) - minimum (red value 1, red value 2)) + (maximum (green value 1, green value 2) - minimum (green value 1, green value 2)) + (maximum (blue value 1, blue value 2) - minimum (blue value 1, blue value 2))
The range for color brightness difference is 125. The range for color difference is 500.

I suspect those formulas for perceived brightness and color difference are for people with normal vision. Accessibility to colorblind people might be a different issue.

When we refer to accessibility, we are referring to the standards and recommendations set forth by the federal government (Section 508) and the Worldwide Web Consortium's Web Content Accessibility Guidelines.

So we'll be "accessible" in the legal sense. Whether we are also accessible in a moral sense depends on how good the standards are for the full range of people. I can't answer that. Show me some web sites that meet those standards, and my daughter will tell you whether she finds them acceptable.

Glenn

[rcrisler1]

The WCAG guidelines are an attempt to normalize visual issues to a standard. It is the only international standard on web accessibility; however imperfect it may be, it's what we've got.

I agree with Angie that we sometimes focus on automatic testing mechanisms; it's easy to test on cynthiasays. We had to build an app, though, to calculate the "goodness" of our contrast values. I'm confident that adhering to those WCAG recommendations will bring us a lot closer to presenting an accessible website for the largest-possible universe of users.

[gledder]

No, that is not all we have got! UNL has some blind and visually impaired students and possibly also faculty and staff. We should be getting their help in assessing the standards.

Standards are only as good as the criteria by which they were established. Does WCAG provide a scientific justification for the standards? If so, that justification should be based on real tests with real people, not on mathematical formulas. I'm not against mathematics--mathematical modeling is what I do. However, I would never recommend that anyone use my model results in practice without doing actual experiments to confirm the results. I can't say that this is a problem here, but I suspect that it is. It is amazing how many accessibility standards are written without sufficient input from those capable of judging. My wheelchair-bound mother occasionally finds that she cannot use a "handicap-accessible" bathroom.

My point from the beginning has been this: Accessibility should be determined by the experiences of people who need the accessibility. We aren't going to get enough visually impaired people to make a real scientific experiment. But anecdotal evidence from a few blind people along with the standards is better than the standards without any evidence.

Glenn

The WCAG guidelines are an attempt to normalize visual issues to a standard. It is the only international standard on web accessibility; however imperfect it may be, it's what we've got.

[rcrisler1]

I don't believe I said that the WCAG standards are ALL we've got. A reader could perceive in your last post an implication that I'm not interested in things other than machine-testable standards; I'm sure that was unintended. For what it's worth, I have had and will continue to have discussions on these issues with relevant personnel at the system and at UNL. There is considerable interest in bringing accessibility to all UNL web pages.

As an international standard, the WCAG is WHAT we've got was what I said. Actually, as an international standard on web accessibility, it's ALL we've got as well, and that is as it should be. It's a strong, well-thought-out set of priorities. At UNL, we may be able to make specific recommendations on the implementation of the standards. It is exceedingly unlikely, though, that we are going to come up with a new set of standards that is unique to UNL. If there are issues that are not addressed in the Web Content Accessibility Guidelines, they should be entered into that set of standards; fortunately, there's an open process to accomplish that.

As the situation currently stands, many sites at UNL are not compliant with even the most minimal accessibility standards, so making them compliant with machine-testable accessibility priorities would be a very significant step along the evolutionary path to a fully-accessible website.

It's critical to users of the web that recognized and ratified international standards on important issues such as web accessibility exist. The current standards were ratified as WCAG 1.0, under the auspices of the Web Accessibility Initiative of the World Wide Web Consortium (W3C), in 1999. That document was edited by two researchers at the University of Wisconsin-Madison, Wendy Chisholm and Gregg Vanderheiden, and Ian Jacobs, the current head of communications at the W3C, the standards body created and currently under the direction of web inventor and MIT researcher Tim Berners-Lee.

A revision of the standards, version 2.0, is currently under development. A last call was recently issued for comments; the comment period will close on June 22. The editors of the WCAG 2.0 draft are Ben Caldwell, Trace Research & Development Center, University of Wisconsin-Madison; Wendy Chisholm, W3C; John Slatin, Accessibility Institute, University of Texas at Austin; and Gregg Vanderheiden, Trace Research & Development Center, University of Wisconsin-Madison.

"The power of the web is in its universality. Access by everyone regardless of disability is an essential aspect."
- Tim Berners-Lee