Web Site Accessibility Blog

The following is an interview with Chris Hofstader. Chris is the former VP at Freedom Scientific and is now actively blogging and enjoying life. Chris’ interview helped to form a lot of the backbone of the article and helped me to understand how an advanced user can use the internet and search engines. Because each of the people I interviewed were so interesting, and they each had their own stories, I thought it best to post some of their interviews for us all to appreciate and learn.
- - -
mb: How long have you been using the internet?

cdh: Since the days before www. Long enough ago to remember when emacs was the best interface to SMTP, NNTP and the little chat programs we used back then. Way before Mosaic or all of these pictures.

mb:What is your search engine preference?

cdh: Depending upon what I am looking for, I usually go to google first but I very much like ask.com and some of the search pages designed for hackers when I need to find a technical tid bit in a hurry. For all “normal” things, it’s google, scholar.google or ask.com.

mb: Is there an accessibility-based reason for that?

cdh: Not really. I’m a power user plus when it comes to JAWS so, no matter what search engine I’m running, I tend to use the JAWS “Virtual Find” function and other tricks to navigate a page quickly. Even really busy pages like ebay and emusic can be tamed very quickly and be accessible to me with a little ingenuity and JAWS features.

Unfortunately, most users don’t know about or care to invest the time in learning how to access such things quickly so struggle with “accessibility.” I think, on search engines, the greatest accessibility problems are no longer issues that can be corrected with the WAI or 508 guidelines but, rather, are design principles for usability. What is “usable” to people with varying levels of vision impairment has hardly been studied and those of us interested in doing so are few and far between so, in my opinion, moving from “accessible” to “usable” will take another decade.

mb: Which engines are more accessible than others, in your opinion?

Some of the oddball hacker ones are pretty inaccessible because nobody stopped to label anything. Otherwise, google and ask both do a decent job and I get where I’m going pretty quickly.

mb: Have search engines ever been an obstacle for you? How so?

Yes. Prior to MSN’s change, it was pretty nasty. Yahoo is too noisy to navigate results efficiently. Google has always been pretty good except for the Turing test thing. I used to like Alta Vista but haven’t looked at it in years. I don’t know a whole lot about others as I don’t even try them.

mb: What things do you do to make search engines easier to use for you?

I use lots of JAWS features for IE. Specifically, I use the QuickKeys, Placemarkers and Virtual Find to move about the pages very quickly. Different strategies work better on some pages than others. Also, I’m pretty good at making logical statements for advanced searches so I tend to get a pretty good hit rate on what I search for. Finally, I have things I search for often stored as favorites, using the facilities for such provided by search engines.

mb: Have they improved in the past few years?

Yes. Ask, google and Yahoo have all improved vis a vis WAI guidelines and such.

mb: Anything else? I know you have some strong opinions . . .

Oh, this is kind of boring. multi-media content can be annoying and, no matter what Bob Regan says, Flash is an accessibility nightmare but I just turn all of that stuff off.

Of course, I search for such geeky stuff that I don’t run into too many multi-media pages. Maybe if I cared at all about POP culture, I might be less of an outlier on such studies. The majority of the stuff I look at is best expressed as text or audio anyway as pictures and such distract from the overt nerdiness of it all.

This is part 2 of my series on Accessible Search. In part one, I posted the full text of the interview with Google’s Dr. T.V. Raman. Using that interview and interviews with end users, I wrote this article, which was posted in short form at Search Engine Watch.

Interestingly, most low-vision or AT users prefer Google or Ask as their primary search engine. Overwhelmingly, it is because of the interface. Both Google and Ask feature a stripped down interface rather than a portal-style interface, such as Yahoo! and MSN. Portal interfaces do not score well with low vision, blind, or AT users because of the “noise” it creates in the experience.
Here is the full article:

The Divergence of Accessible Search

Earlier this summer, Google released an Accessible Search Beta through its labs projects. The Accessible Search was the personal project of Dr. T.V. Raman, who came to Google about a year prior. The move by Google to release the project came on the heels of Google changing from the visual-based security device of the CAPTCHA, where users have to type the letters that they see in an image. Google added an audio component as a response to online petitions and outcry from the visually-impaired users. Blind users were not able to access many Google services, such as GMail, Blogger, Google Groups, and a Google Account as the visual-based security kept them from participating in Google’s services.

Dr. Raman comes from a background of advocating standards-based programming and structured data on the web. This made its way to the Accessible Search algorithm that favors standards-based websites. This was a natural method of creating the accessible focus of the search:

“For accessibility, clean, well-structured semantic markup is the best thing you could hope to have. The three key phrases “clean” “well-structured” and “semantic” are all important in the above. Adherence to specs definitely makes checking easier. . . I believe that Web sites creating clean, well-formed XHTML content will, over time, find it significantly easier to serve all their users better — simply because their content will be easier to manage and evolve.”

The Accessible search was received positively by the Blind community, yet many low-vision users took issue with the tagline of “Search for the Visually Impaired.” This was typical of larger issues in the accessibility community that many sites are accessible for blind users, yet the same amount of accessibility is not available to low-vision users, who require different methods in order to access the content.

Regardless of a few criticisms, the reception was positive overall. The end result was that users found this to be a useful tool. “And that’s our goal,” said Dr. Raman.

Despite the contention among the different groups, Google is widely accepted as the search engine of choice. The main reason is the uncluttered interface, which makes it easy for users using magnification interfaces, screen readers, or a joystick interface.

Diversity of Access

The main obstacle is that accessibility is varied by need, rather than a single fix. Because of the varied nature of access, assistive technology, and multiple needs, sites must cater to a host of diverse combinations.

Here is a small sample of that diverse group and how they rely on the search interface:

Lee

Lee has albinism, and one of the main factors is low-vision. Lee is a long-time computer and Internet user and has had extremely low-vision since birth. The growth of the Internet has been the opening of a limitless world for her, as phone books, maps and other small-print formats have all been replaced with internet versions that are easier to understand and read.

She relies on the Apple’s built-in magnification software in her daily browsing. She prefers to view yellow or white text on a black background for added contrast. She prefers to use the search engine Big.com, which is powered by Ask. Big.com is built specifically for low-vision users, enabling them to read the large-font results without eye strain.

Lee explains how she sees:

“I have a difficulty seeing details at a distance. I also have problems reading small print and light and glare are an issue as well. I have nystagmus (an uncontrollable and irregular motion of the eyes). The world doesn’t move, but my brain has to work harder to provide a stable image. Eye fatigue is an issue. Compared to most people with albinism, I have fairly good vision and am able to get by with minor modifications.”

The Accessibility search was not helpful to her, as it was obvious that it was initially created for blind users. However, she is a Google Search fan. She would like the ability to store her Google preferences through a cookie so that her background and text color preferences would be automatic.

Lee is very active in the albinism community, which shares information, but she was not aware of the new Accessible Search feature. She feels that search engines respond to the accessibility community very slowly and thinks that information passes slowly as a “trickle-down” effect.

Glenda

Glenda has an inspirational story and is currently writing a book about overcoming Cerebral Palsy and building a life for herself. Glenda uses a combination of assistive technologies to interact online. Her main input device is a keyboard, on a non-slip mat. This is critical as she slides her hand along the top of the keyboard and uses her left thumb to hit each key. Any movement or involuntary jerk would move the keyboard if not for the non-slip mat. This also affects her use of a standard mouse. It its place she uses a joystick to control the cursor movements. Along with this, she uses word prediction software that predicts the words that she is typing, which saves on keystrokes.

Glenda does not have any vision impairments, but she relies on interfaces that are clear, simple, and have a consistent design. The consistent design makes it easier for her to find information, while cluttered pages and flashing animation distract her from easily navigating the page. She prefers the Google Toolbar because of the simple interface and availability for searches.

Even though she is a ten-year internet user, she relies on simple interfaces to use websites, read pages, and interact. She does not need the low-vision or blind accessibility features, but rather features than enable her to input information easily and do not distract her by moving or flashing elements.

Chris

Chris is a JAWS power-user; in fact, he had a major hand in its development. He is blind, but is able to use the internet and related technologies faster than most sighted persons. He has been online long before HTML was the protocol and used to use emacs as an interface to a chat program.

Chris prefers Google and Ask, mainly for the clean interfaces they offer, but also likes using search pages designed for hackers, especially when searching for technical information. No matter which search engine he uses, he is able to use all of the tricks in JAWS to filter out the noise and find specific information.

For users who don’t understand how to use JAWS, Chris understands their frustration:

“Unfortunately, most users don’t know about or care to invest the time in learning how to access such things quickly, so [they] struggle with “accessibility.” I think, on search engines, the greatest accessibility problems are no longer issues that can be corrected with the WAI or 508 guidelines but, rather, are design principles for usability. What is “usable” to people with varying levels of vision impairment has hardly been studied and those of us interested in doing so are few and far between. So, in my opinion, moving from “accessible” to “usable” will take another decade.”

To Chris, and many others, it is a factor of usability, which is why Google and Ask get high marks for their simple interface. Yahoo and MSN search are not rated highly because of the portal style interface that is loaded with information and links, which is “too noisy”. Search engines that use the Turing test to filter bots are also a huge obstacle to enjoying many of the custom features. Google’s removal of the Turing test was big news to the blind community this year.

Overall, Chris feels that search engines have improved as a whole, mainly by integrating more accessibility and standards-based programming into their results pages. But no matter what anyone says, he still finds Flash to be an “accessibility nightmare.” He seeks information that “is best expressed as text or audio.”

Continued Development at Google
Obviously, progress hasn’t stopped for future development of Google’s accessible search. Recently, the Accessible Search has been improved to give blind users access to the same advanced search features that are available in the regular Google search.

Dr. Raman, who has been intimately involved with building speech-enabled interfaces for many years, sees the Google Accessible Search as “a good first step in delivering the most accessible information online” and that Google will continue to develop ways to be more accessible.

He welcomes any user feedback, as “we get better when we know what works for the user and welcome feedback on everything from improving taglines to adding new features.”

Dr, Raman feels that the impact of the Accessible Search will go beyond a simple service and one that will affect the development of the web. “I hope that as we continue to improve Accessible Search, we’ll have a long-term positive impact on increasing awareness of the need to create clean, accessible Web content.”

This is part one of an interview series that I had performed this fall while researching for an article. The article ended up being much too large for publication, so I have decided to publish the interviews and the full text of the article here at Accessibility Blog.

Part one is the interview with T.V. Raman, who was instrumental in creating the Accessible Search interface at Google.

How was the idea for the accessible search received?

Working on Accessible Search has been a very positive experience for everyone involved. From the start, within Google, the idea of Accessible Search was received as a useful tool and an important first step. Interestingly, although there was a lot of feedback from the accessibility community as to how Google could improve accessibility to some of our services, the actual idea of Accessible Search as a tool to help find accessible content wasn’t necessarily on top of anyone’s list. However, from our perspective, we thought that offering new search functionality built on what we already have in place, was the fastest and best route to helping users discover pages that would be most useful to them.

I think our instinct was right because Accessible Search has been a very positive experience from the start and continues to be so. That goes from building it to seeing the initial positive reaction from the user community and continuing to improve the product to date. I hope that as we continue to improve Accessible Search, we’ll have a long-term positive impact on increasing awareness of the need to create clean, accessible Web content.

It seems as though you are taking into account some of the elements of standards-based programming in the algorithm. Are standards anything of concern at Google? Do you find that sites written according to W3C standards are more relevant in general?

For accessibility, clean, well-structured semantic markup is the best thing you could hope to have. The three key phrases “clean” “well-structured” and “semantic” are all important in the above. Adherence to specs definitely makes checking easier. However, I do not believe that passing a checker that validates ones markup for structure is a sufficient condition to verify accessibility. I can easily create XHTML documents that will pass every static checker but may potentially end up not being accessible, i.e. any one measure is very easy to game. That being said, I believe that Web sites creating clean, well-formed XHTML content will, over time, find it significantly easier to serve all their users better — simply because their content will be easier to manage and evolve.

Do you have ways of distinguishing appropriate and inappropriate ALT attributes?

Google scores ALT attributes appropriately, and things like stuffing a 14-page essay into an ALT tag won’t help you rank better and will annoy your users. We encourage users to use ALT and TITLE tags in the best way for users, and that work well for Google as well.”

In general, what issues does Google believe it can authoritatively decide are “accessible” - given that only the validation of code and absence of deprecated features can be fully tested automatically?

I believe our “abilities” in the above are likely to be a moving target. I think of it as being analogous to our ability to detecting email spam. Also, keep in mind that we’re still soliciting feedback from the user community and evolving it further, which is why it was launched in Google Labs.

How has the accessible search been received by blind users?

So far, its reception has been very positive. There’s been some criticism in the blogosphere saying “this isn’t needed, sites should just follow access standards” but end-users who are struggling with the mess we have on the Web today definitely appear to find this a useful tool and that’s our goal.

Is Google working with the blind community to make them aware of the Accessible Search?

We’ve announced it on many of the blindness related lists and have spent time briefing some of the journalists that cover this topic. In general, the blind/visually impaired user community in the U.S. is very well connected, in that users become aware of services fairly quickly. If it’s useful, word gets around quickly. We’ve relied on word-of-mouth, just like we do for so many of our product launches, which is a testament to our usability.

Many low-vision users weren’t impressed, as the tagline said for the visually impaired, but wasn’t any better than the usual search. Are there plans to expand the accessible search for all vision types?

Let’s separate the tagline from actual user experience. We got a lot of flak from different users just on word choice e.g. “visually impaired” vs. “visually disabled” vs. “visually challenged” vs. …and, personally, I believe it’s impossible to come up with a catch phrase that covers everyone, makes everyone happy and at the same time conveys the entire gamut of the user population and the exact meaning of the product.

As for search results with Accessible Search, our intention for this Google Labs experiment was to see if we could reorder results based on accessibility. It seems that many users just saw the name “Accessible Search” — and then went to the results page expecting an entirely different layout of results. When they didn’t see something entirely different, then went away thinking it saying “it just looks the same.”

Since more than one person had this experience, our take away is that we need to get better at communicating what Accessible Search is really all about. We get better when we know what works for the user and welcome user feedback on everything from improving taglines to adding new features.

Are there more plans for Google to become more accessible?

Building Accessible Search is a good first step in delivering the most accessible information online and we hope to continue developing ways that Google can be more accessible.

Are you working on additional accessibility projects?

While we don’t talk about specific products before they launch, I’ve been thinking about a few projects in this area.

Can you give us a little history about yourself?

Absoutely. http://emacspeak.sf.net/raman