The Guild of Accessible Web Designers | Don't use the statistics defence as a reason to exclude people from your content

Don't use the statistics defence as a reason to exclude people from your content

Photo: Jim Byrne. Contributed by Jim Byrne

I was recently involved in a discussion about whether website designers should be expected to accommodate Netscape 4 users.

The case against accommodating Netscape 4 users is invariably backed up with statistics about how few people now use this, admittedly flawed, browser. I've heard 'the statistics defence' (as I will call it) so often over the years that this latest evocation prompted me to think about why I don't agree with this approach.

My thoughts and arguments against the statistics defence are not yet fully formed. I would welcome any feedback on the subject. It is such a common argument against accessible web design in general, that a page containing counter arguments would be a good resource for web accessibility advocates.

Examples of the 'statistics defence':

"We design for 17" screens because that's what most people use these days"
"We assume 92dpi resolutions because most people use a PC"
"We use IE 5 as a baseline because very few people use old browsers now."
"We don't provide an alternative to our flash site, because everyone has the flash plugin these days."
"We don't need to make our site accessible because it isn't aimed at, and doesn't get used by disabled people."
"We design our site to work on 600 * 800 because that's what most people use."

My arguments against this approach

I'll give my conclusion first: content on web pages needs to be accessible to Netscape 4 users - and all the other user agents accessing web content. The argument that we can ignore a particular set of users - because they only make up a small percentage of our audience (i.e. they use a particular browser or a particular bit of access technology) - isn't one web designers should be buying into. It is irrelevant whether a person is using Netscape 4, a screen reader, or a keyboard driven text only browser - the issues are basically the same; it is about accessibility of web content.

The statistics defence assumes users needs and user agents are predictable

What assumptions do many web designers make about their intended audience. e.g. what browsers do they assume they are using? what Screen size? screen resolutions, bandwidth, colour pallette? Are those assumptions based on the computer they have on their own desk, i.e., the one they are using to design the website? Probably - but is this a good approach? - probably not.

Have any of the following things changed in the past: browsers, hardware devices connected to the web, screen size, screen resolution, Markup versions? Of course they all have. Will these things change in the future? Yes - all of them. Designing for a specific configuration of hardware and software isn't a good way of making pages future proof. Even users with the same hardware and software resize their browser windows to suit their own preferences.

A vital lesson to learn is - change is the norm: the most predictable thing we can say is that everything changes. The best chance we have of dealing with this unpredictability is:

Use standards so that sites have the best chance of working on the widest range of user agents.
Create sites that are flexible enough to deliver our content - no matter what the end user is using.

That is not to say that the presentation will be the same on every device - it won't be. The presentation is important - but if the content isn't accessible - the presentation doesn't matter - because there is nothing to present.

The web isn't paper

Cross platform/cross browser compatibility is the strength of the web - that was the problem it was designed to solve. Designing a web page is not like designing an advert or a bus shelter or a magazine page or a document to be printed on a sheet of A4; where the amount of 'real estate', colours, text size and so on is predictable.

To take the specific issue of access for disabled people; do we have to accommodate the needs of disabled people? Do we have perfect knowledge about their access needs? The answer to the first question is yes; in the UK, the Disability Discrimination Act tells us that we can't discriminate against disabled people. The answer to the second question is no; we don't have perfect knowledge about the access needs of disabled people.

10% - 20% of people in most populations have some kind of impairment: some of those impairments are not obvious: 8% of men have colour blindness (.4% women) - approx 5% pop with visual impairments - approx 5 - 15% Dyslexia. Once people get older (say over 40) their eyesight, hearing and motor skill start to deteriorate

In the university where I work we have many disabled students - not all of them are registered as disabled, but approximately 500 are.

Impairment	Approximate Numbers
Dyslexia	230
Blind/partially sighted	24
Deaf/partial hearing	25
Wheelchair.mobility	21
Autistic or Asperger	2
Mental health	10
Unseen disability (Epilepsy, diabetic,etc)	91
Disability not listed	101
Two or more of the above	21

We don't have perfect knowledge about the access needs of each individual listed above - so we need general strategies to deal with this unpredictability. In terms of approach, dealing with the diverse needs of disabled students isn't much different from dealing with the problem of making sites work on different browsers and different hardware platforms.

We have to assume that we don't know what the end user will be using - or what their access requirements will be - and think about what this when we make design decisions. If it turns out that our content isn't accessible on a particular browser - we need to find a workaround to solve the issue (while maintaining standards markup and accessible design). There is always an answer - even if sometimes it take a bit of time to find it.

We have to make our websites accessible because it is the law (in many countries).

In the UK we have the Disability Discrimination Act and the Special Needs and Disability Rights Act: and in a university that means we can't discriminate against a student on the grounds of their impairement; reasonable adjustment and anticipation of students needs is required.

We can't argue that we won't accommodate disabled students because they only make up a small percentage of the student population. Equally we shouldn't argue that we won't accommodate users with particular browsers because they are part of a minority. In relation to the particular case of Netscape 4, it is legitimate to ask users to upgrade so that they get both the content and the good design - but not legitimate to argue that they won't get the content if they don't upgrade.

A note for those who have not understood my point (i.e., as is clear from the comments below)

I am not advocating support for Netscape 4, I'm advocating support for accessible content. In practice this may mean serving Netscape 4 users content/markup but without a style sheet - so that the still get your content but don't get a broken visual design - which in practice could make the content inaccessible.

Comments

It is simply not possible to make your website 100% accessible to everyone. However it is always a good way to improve accessibility of your website day by day.

Mike | Mon Apr 13 2009

I agree that there should be a baseline standard which we all adhere to, but we cant consider every medium / browser as necessary to support forever, else we will eventually dwarf our own potential for improvement. For example I use flash a lot, which has many accessability issues, but still has merits and uses. If accessability was essential for 100% of people then this technology may never have come into existence

Flash Brighton | Sat Apr 11 2009

I might not have correctly understood your point but statistics has great impact on your decisions to make website development strategies for a successful website launch.

Ashley | Thu Mar 26 2009

I am a strong believer of Change is the norm. Usually we forgot about users using older technologies and hence just move forward along the road of new technology inventions and improvements. There is no doubt, this causes accessibility problems for many users. A base line should be defined to make your website accessible to all users no matter if they are using older technologies or disable person.

Urdu | Thu Mar 19 2009

I agree that it is important to make websites accessible at a base level, but it is surely prudent for web designers to geer their designs towards the best available technologies. This may mean that eventually it is necessary to stop supporting certain browsers which become outdated. Otherwise how will technology and design on the web evolve, if we have to drag along outdated techologies forever? Saying this adopting principles of progressive enhancement and using css properly should allow graceful degredation anyway,in nearly all cases.

bozboz web design | Sat Nov 15 2008

Statistics make 80% of the world go round. Jim, I have to say that I don't think you can have it both ways. It would be difficult to justify not taking notice for statistics about browser use, screen size etc. but taking into account statistics about disabilities.

Richard Morton | Tue Sep 02 2008

Interesting point from Ken about the analogue TV switchover. It is true that there is a graceful degrade path created by the introduction of set-top boxes but of course graceful degradation doesn't go on indefinitely. I'm old enough to remember the switchover from 405 line to 625 line TV and their was a graceful degradation then too, with TV's switchable between the two modes available well before the switchover, but if you kept your old 405 line only TV it wouldn't still be usable. If you've bought an HD ready digital TV, you might well think that means you will be able to watch HD freeview broadcasts when they come out - think again, in most cases the HD ready part means just the DVD etc. In other words it isn't possible to cater for every possible combination of user setup.

Richard Morton | Tue Sep 02 2008

"Hi Keith, I'm not saying anyone should use Netscape 4 - that is not the point of the article." "The case against accommodating Netscape 4 users is invariably backed up with statistics about how few people now use this, admittedly flawed, browser. I've heard 'the statistics defence' (as I will call it) so often over the years that this latest evocation prompted me to think about why I don't agree with this approach." If you disagree with the statistics approach, you really should have directly all of your energy at that instead of ever mentioning Netscape 4, because that's just a horrible, horrible example.

Paul | Mon Apr 14 2008

Came across this while looking for web-stats. I read the first paragraph "My arguments against this approach" but didn't find any arguments in the paragraph, so I stopped reading. I may read further when I have more time but to have an opener titled "My arguments against this approach" which contains simply statements of belief, with no reasons supplied, implies the rest of this document isn't going to be worthwhile. I wonder how many others have gone elsewhere after reading that. Please amend the first paragraph to at least provide a summary of your reasons.

tom b | Fri May 18 2007

Jim, People still use Model Ts but we don't build our road infrastructure around them. We just allow our roads to "degrade" gracefully so that Model Ts can get from A to B. Look it is a business and function decision. If you are targeting a website to enterbrise business users in the USA, then IE 6, IE 7, Opera, Firefox, Safari and Konqueror and you are there. However, if you are a government department, giving out public information, you better be sure that the text gets through on just about everything. Another example: We are about to shut off analog TV in a few years despite a huge installed base. The "degrade path" is to get a digital tuner box that converts and outputs "old" formats. Inconvenient for the analog user but it will work. "You can ignore old browsers" is not a hard and fast rule that shold be followed, but neither is "support all browsers with all of their quirks".

Ken | Mon Dec 11 2006

Okay, I see a lot of people weighing how to feed the hungry of cyber space, but in a non-academic business world, things work a little differently. I spent the last three weeks working close to 90 hours per week as a web team of one for a well-known corporation. There planning was beyond piss poor, but the goal was clear: deliver it on time. Great! Where are the other seven developers? Where's the f-specs, prototyping? Have we asked critical development questions? What are we supporting? Have we looked at the stats? Stats? The simple answer is hell dam no. So, I do the impossible and build the beast to work in Firefox and IE, 95% of our users according to the logs. Day 1: Employee tells me it's not working his cell phone I explain, we don't support PDAs, as it's less than 1% of our user base. Dickhead, emails our IT guy (of course, should've seen weeny boy coming down the hall). IT guy points out that wingnut OSs constitute 10% of our traffic. I point out that the most popular browsers for those wingnut OSs constitute the 1-2% crowd (suck my left one). So now, I'm sitting wondering if I'll ever have a weekend again, needing the dough all too badly and faced with the monsterous task of supporting browsers that nobody uses. NOBODY USES THEM. BUT alas, my response is simple: Mr.Bossman, even as a salaried employee working 24 hours per day, it would take 10 weeks of production away from the the other ten projects you'd like to push out the door, and all to satisfy some guys that may or may not be our customers? Does this sound like a sound bottom line proposition to you? Because what it really means is that if you want more, you pay more, i.e. bossman would need expensive contractors to make the site compatible with every blind persons needs in under three weeks ... and lets make sure the cat and dog can use the site too. No. Outside of the slack of academia, there is no room for pleasing all the people all the time, and if there were, there'd be no toilet paper in academia. Ugh. I hate alternate browsers, almost as much as I hate IE.

Brent Forbes | Mon Feb 13 2006

The Netscape 4 argument is really one that is dead. Simply use sound and modern xhtml table-less web design practices. Assign either a media tag to the CSS style sheet link properties or use the @import rule so N4 and most other unruly old and tired browsers just bypass the CSS completely. Making a website accessible does NOT mean catering to all outdated equipment and software so they LOOK great. It simply infers that the content can be read and navigated by these users. For those who beat this N4 horse up continuously, you have a whole lot to learn about good xhtml/css practices. Today's designers and developers alike have the new N4 to deal with, namely I.E. and its poor CSS 2 support along with its countless security holes.

Rob | Tue May 03 2005

These are my personal feelings on the subject: There is no defense for web inaccessibility. I say we, as accessible web designers/developers, make the web accessible to everyone. We exclude no one. That is the whole point. We set a good example to others. If someone takes the time to come to my website (one that I have built), by golly, I want to ensure they can get the goods - even if it's in black and white, even if those goods are delivered as plain old html, or even text or sound. If I use technologies unsupported by someone's access device, which is to be expected nowadays, whether the technology is Flash, CSS, descriptive imagery, or JavaScript, I have to ensure there is another way to make the content accessible. This, however, is not that hard. After all, this thinking ensures cell phone users can get the goods as well, and there is nothing antiquated about that. It's not that difficult. Start with the html and build from there. The rest is simply eye-candy. All I have to do is make an accessible link to the content. I try to empathize with the handicapped person, whether the handicap is physical, mental, or technological, as they have a right to my content if they've made it as far as the internet to begin with and have an interest in my subject matter. As far as I know, the only impossibility is to make the content understandable to someone who cannot understand. I try to write clearly, but if it's a deep subject, what can I do? I cannot make it un-deep. Statistics have been, and always will be, cited, used, and/or manipulated by someone so I can't take that much stock in them. Sincerely, Mike Cherim

Mike Cherim | Sat Mar 26 2005

With respect, Roger, the Web isn't a car or a mansion. It's a medium that is more than able to encompass everyone, regardless of their abilities. Accessibility isn't about flaming web developers; it's about getting web developers to focus on their audience, rather than their egos. Guidelines are in place for user agent manufacturers, but they're useless if web developers don't at least take on board the concept of making their resources available to all. I thought the myth about accessible websites being ugly was well behind us now. The appearance of the website will depend on the talents of the designer. There are no constrains that require that accessible websites are ugly. The resulting website will depend on the talents of the designer, not whether it is accessible or not. [The] majority of sites are for [b]usiness, and have little concern for the disadvantaged. That[']s business! A good business model would not exclude part of its audience. It makes sense for a business to ensure that people with disabilities can use services, as it will increase their revenue. Services offered online have the advantage that there are no extra costs, although obviously it would require the developers of being capable of doing their job properly. Put your prejudices aside, and increase your target audience. That's business!

Gez | Sat Mar 26 2005

I really do apoligise in advancee for any offence thius causes! I don't see cars designed for blind people.....I don't see p[oor people living in mansions.....I don't see people with aids being treated to a full life! Is that discrimination? No, it's quitre damn residual and obvious and to the affect of causation! Sorry, yet when people start getting on high horses- claiming to champion for fellow man, ~I get nervous... normally it morales an ethics mixed with ego. I.T. SHOULD be available for the disadvantaged, yet there is no way to build the majority of sites so that this is possible. You claim that you can be sites that degrade nicely, particularly using software... yeah, and they look pants, and most sites end up looking the same! MAYBE it should be the BROWSER designers and SCREEN READER designers that need to pull their socks up, and not flaming web designers. A majority of sites are for pbusiness, and have little concern for the disadvantaged. Thats business!

Roger | Sat Mar 26 2005

While you all make valid arguments on both sides, saying that all sites should be made a accessible is like saying that all print material should be in black in white because there are those who are color blind. Let's face it; you can't please everyone, why not go for the majority.

Ken | Thu Mar 24 2005

come to this thread a bit late I know but endulge me please. the company i work for stopped supporting NS 4.x when it moved to a css/xhtml format for all its sites. it was a decision made for business and workload reasons, but justified by the statistics arguement that Jim has highlighted. at the time i had no problem with it, it saved us a load of time and weeks worth of css hassles. it was a hard enough job trying to convince some of our designers about the merits of CSS without having to explain why the NS 4.x looked so different. however, some 2+ yrs on now, i am finding myself drawn more and more to the challenge of making it work. i understand there are technical reasons for trying to avoid working with these browsers when using css, but morally we should be sending the pages to these 'legacy' devices with as much care and attention as we send them to the more modern ones.

Ian Burrett | Tue Dec 21 2004

Jim As a student studing web certificate with the Open University I was quite interested to read your comments about the need to develop for Netscape 4 browsers. Part of the work we've been asked to do, is to look at designing a web site for a theoretical company. Having gone through the course, and been told how good designs should make use of CSS in order to place objects on the screen rather than making use of tables - I have been having all sorts of problems using CSS with Netscape 4 I don't know why, but it seems as though I can position up to about 4 'id' tags on a page, any more and NEtscape just goes mad! Any help you can offer in order to make my site work with Netscape 4 would be appreciated - as I agree although the stats talk about 0.2% of the population make use of this browser as part of my assignment Many thanks Andy

Andy Knott | Fri Dec 10 2004

Excluding Netscape 4 is an educated decision, and a move in the right direction. We all know that we have to build site to suite our users; they run the show. But it is self destructive to develop for users who are uneducated, and are using outdated, invalid browsers such as Netscape 4. By sacrificing this browser on the altar of a better experience, you could help people to wake up and realize the tools they are using are not the right ones. Its not their fault they are uneducated, its yours. Developing for browsers like Netscape 4, is the same as saying that you SUPPORT netscape 4 as a browser for internet users. It is a hard decision to make but at this point in the game, the statistics show that its time this browser got the cut.

Eddie Wilson | Tue Nov 02 2004

Hi Keith, I'm not saying anyone should use Netscape 4 - that is not the point of the article. All the best, Jim

Jim Byrne | Mon Oct 25 2004

Is there actually any good reason for anyone to be using Netscape 4 these days? Blind people use screen readers because they have other choice. As another obvious example, Linux users do not have the option of installing Internet Explorer to view pages which only display properly in IE. However, anyone using Netscape 4 is presumably using a computer capable of running a more recent, and far more standards-compliant, browser such as Opera or Firefox. So why would they choose to stick with an outdated and notoriously faulty piece of software which must make at least 50% of websites look like the proverbial dogs dinner?

keith westminster | Mon Oct 25 2004

When we sell our web services, we do not rely on the accessibility angle too much. We find very little (no) resistance when we explain that we just build things properly - then show them what a print stylesheet can do - and they are sold. Accessibility becomes a side effect in the clients/payers mind. Good will to fellow man and business bottom line unfortunately don't live well together. I like the fact that business gets a site that does all the things they want (SEO, attractive, quick, usable) while we know that what we have done also includes WAI measures. Instead of those (who are not on the standards train) asking the rest of us "Why?" and "Justify" - it is us who are asking them "Why Not?". Why are you building sites that don't allow most people to absorb it? It does not have to take any longer to make something accessible, so why not do it? why does you site print-out look like mashed vegetables? why does your code weigh 20Kb when ours comes in at 8Kb? Why are you bothering with this industry? Why are you developing sites with obsolete knowledge and technology? Some of that is too arrogant, sure, but you get the point. There is a right way of doing it (name all bonuses) and there is a not so right way (name any benefits?) It does not need to be a 'help those less fortunate' argument. It is a "Just do it properly" argument. apologies if i am not really adding anything to the argument.

Dave Rayner | Wed Oct 20 2004

There are 15,000 students - so 500 is just above 3% of the student population. This is an underestimate of the number of disabled students in the university, as it only includes those who have chosen to register as disabled. Jim

Jim Byrne | Wed Oct 20 2004

Jim: What percentage does the 500 make of the total number of students at the university? Jules

Julian Rickards | Wed Oct 20 2004

Hi Mark, Unfortunately I don't know of sources of statistics related to the number of disabled people who use the web. However, you may be surprised to know about the statistics compiled by the site ethnologue.com: "Blindness and Braille. There are reported to be from 23,000,000 to 40,000,000 or more blind people in the world. " http://www.ethnologue.com/ethno_docs/introduction.asp

Jim Byrne | Tue Oct 19 2004

To me, often a word that comes to mind about this stuff is this: discrimination. It is discriminatory to cut out people, even based on statistics. Sure, there will be some gray areas, but if your philosophical approach to web development is inclusion rather than discrimination and exclusion, then I think this should not be a problem. Tools exist to EASILY create sites that will degrade, in many cases. Sure enough some sites will be hard to make accessible (a picture gallery, or some animation or video where transcripts are not available or legally possible etc etc), but on the whole discrimination can be avoided quite easily.

Bob | Tue Oct 19 2004

I agree whole heartedly. I make sure the content makes sense when read as plain text, then I incorporate the design for most browsers. BTW, where are some sources for detailed statistics on disabilities? I often get bombarded with the "how many blind people use the web" statistic defense. I've done a cursory search, but either they are incomplete, hard to get or hard to interpret.

Mark Darbyshire | Tue Oct 19 2004

Related: