Why Inequalities Can Appear In Digital Accessibility and Inclusive Design

The world is full of questionable design outcomes due to unconscious biases, political agendas or a lack of empathy, and working as a manager in accessibility you have to be mindful of all of these in your research and decision making processes. These are not just things you have to deal with but you also have to be mindful of not letting them creep into your own practice.

Taking the whole audience on an accessibility journey is a difficult thing to do. Accessibility has many perspectives and because of the history of accessibility so far, for different groups of users the accessibility model is maturer than it is for others. Not everyone has an equal voice or in some cases even a seat at the table. This can create an imbalance in discussions, oversimplification of complex problems, intersectionality being forgotten and a skewed perspective of what inclusive design and accessibility actually are.

Quite rightly we use demographics in our framing of any accessibility opportunity. We talk about the 20% of ‘disabled’ people in the United Kingdom of working age. We can add to that numbers of people not included in the definition of ‘disabled’ by the UK government, including people who are dyslexic or dyspraxic, those who have ADHD, those who are unaware of their hidden condition but still experience barriers, people who have impairments or conditions but do not identify as “disabled”. All of these people still experience barriers and should all be considered.

I have been in many meetings where we talk about the big data, but the suggested approach is only impactful for a subset of that audience, which if not flagged creates inequality within equality.

This is apparent in the regulatory frameworks. The big statistics are used but the laws and guidelines are still rather selective with many people included in those stats being left out of the solution. Their headcount is beneficial to the argument but the spoils are unevenly shared.

This is not always avoidable because the that is the way accessibility has evolved. WCAG for instance talks about disabled people collectively and yet it is still very much a work in progress from several perspectives. A good example are the guidelines on closed captioning, the user experience elements of the guidelines are still at least 20 years behind broadcast TV in the UK, and any guidance on video production has still not yet entered this century. The W3C’s cognitive accessibility task force(COGA) is a fantastic but relatively fledgeling group helping the guidelines play catch-up from a neurodivergence perspective, etc.

The potential pitfall with using WCAG in isolation is in expecting it to contain all the answers for everyone. It is worth remembering that guidelines are a resource and starting point, but are never a complete answer to accessibility because they are a work in progress and have unintended certain biases created by their process of evolution. Two of the main biases in WCAG are socio-economics and politics. The groups that write the guidelines are full of clever and insightful people, but there are a limited number of perspectives. Disabled groups are not evenly or proportionally represented. Some groups have no political channel for representation, they have no voice as it were, and still have no seat at the table. A good example is people who actually have no voice or have language barriers, and therefore no autonomy. The working groups are full of middle class professionals, like me. We all have access to the latest platforms and technologies and have only our own lived experiences to draw from.

None of this is intentional bias, but a privileged perspective of the creation of any guidelines is hard to avoid when the overwhelming majority of people who write them are in some ways privileged.

The point I am not making here is to not use WCAG, quite the contrary, please use it, it’s great, but it is an incomplete jigsaw because not all of the pieces have been found yet. So as an accessibility professional it is important to be mindful of the gaps and work with your design and product teams to address them where you can.

For the rest of this article I am going to break down three example projects from my time at the BBC to show where there are other lenses to accessibility beyond guidelines.

These examples will focus on the socio-economic lens of accessibility and how we can often forget that not everyone has access to the latest platforms, broadband, assistive technologies, or because they simply have no way or being heard..

In the UK socio-economic accessibility affects upwards of 5% of the population. They often have no easy channel of communication and as such are all too easily forgotten.

The BBC is a global brand with an audience in the hundreds of millions, and in many of the regions that it operates, the percentage of people that have both accessibility and socio-economic considerations, the percentage of people affected can be considerably higher.

BBC World Service (WS) web and mobile app services provide regional and global impartial news and analysis in 44 languages and its digital content is accessed by over 500,000,000 people every month.

The weekly trackable audience to the main World Service website alone is around 100,000,000 browsers;

· 85% on mobile phones (smart and feature)

· 50% of the audience under 35

· 42% female to 58% male gender split

In Afghanistan, Somalia and Myanmar, World Service is used by up to 30% of the internet population.

As well as the website, the World Service supports other digital distribution methods:

· Syndication, so content appears on tens of thousands of regional websites and social media accounts.

· Accessible downloadable formats are distributed by USB, CD, torrenting and other file transfer sites.

· They utilise Telegrams as a format in many regions and have a mirror site on the Dark Web that provides news in regions where BBC services are blocked for political reasons.

· BBC WS Radio app has a ‘dial-to-listen’ feature in regions where the cost of a phone call is cheaper than streaming data.

For the BBC World Service, accessibility is as much about socio-economics barriers as it is about physical or cognitive impairment. In many of their regions end user equipment can be considerably older and expensive assistive technology software licences or physical technologies simply aren’t available.

So the next time you look at the WebAim screen reader survey be mindful of where the responses came from and you will realise that that it presents a rather privileged perspective of assistive technology usage.

Disabled people in lower or middle income nations, regimes or in war zones do not have access to high-end technologies, affordable data tariffs, safe and secure accommodation, or reliable unbiased information.

The BBC World Service design their websites from technical, functional and socio-economic accessibility perspectives. Their approach supports a more realistic view of available technologies and platforms including a huge range of legacy devices and small screen sizes. This ensures browsers like Opera Mini are high on the priority list (53% of the Hausa language service in 2020).

For blind, vision impaired and other screen reader users there is a problem that is out of the BBC’s control, that out of the 44 languages that the BBC provides news services in, between Google, Microsoft and NVDA, only 34 are made available as screen reader languages. The websites are future proofed for all 44 languages when and if screen reader providers are able to extend their language provision. There is a bias again as language support is aligned more with how wealthy the nations are so in Africa for example the languages supported are either European (English, German, French or Dutch) or Afrikaans.

These are the current screen reader supported languages supported by the BBC World Service:

Arabic, Bengali (Bangladesh and India), Burmese, Chinese (UK, China, Taiwan and Hong Kong), English (UK and India), French, Gaelic, Gujarati, Hindi, Japanese, Korean, Kyrgyz, Marathi, Nepali, Polish, Portuguese (Brazil and Portugal), Punjabi, Russian, Serbian, Sinhala, Spanish (Colombia and Spain), Thai, Telugu, Turkish, Ukrainian, Urdu and Vietnamese.

When considering people’s economic and political challenges the websites are also able to gracefully degrade or adapt their user experience without dumbing down content. No JavaScript experiences and very light page weights have to be of paramount importance, because as well as there being uncertainty about end user equipment capability, bandwidth can be limited and data can be astronomically expensive when compared to living wages.

Additional richness of user experience of experience come from a lighter version of the BBC Media Player and the provision of open video captioning using BBC Reith/Qalam to improve readability and legibility. Instead of audio description being used all reportage is predominantly produced as if it has no images, so it is accessible in it’s default state and can also be reused by one of the many BBC World Service radio stations. The core of the web service is focused on text-based content because of its universality. Accessing video is still a major socio-economic issue in many of the regions, which is why the mobile app provides a dial-to-listen service for its audio streaming services in the worst affected regions. Phone calls are a lot cheaper than streaming audio.

This is about accessibility empowering people through equal access to information which is a key deliverable in The United Nations Convention on The Rights of Disabled People.

In Article 21 — Freedom of expression and opinion, and access to information:

States Parties shall take all appropriate measures to ensure that persons with disabilities can exercise the right to freedom of expression and opinion, including the freedom to seek, receive and impart information and ideas on an equal basis with others and through all forms of communication of their choice, as defined in article 2 of the present Convention, including by:

a) Providing information intended for the general public to persons with disabilities in accessible formats and technologies appropriate to different kinds of disabilities in a timely manner and without additional cost.

d) Encouraging the mass media, including providers of information through the Internet, to make their services accessible to persons with disabilities.

BBC World Service gives the opportunity of equal access to information to millions of disabled and other disenfranchised people in some of the hardest to reach regions. It is able to do this because their product teams and accessibility lead took the time to understand where the barriers are and benchmarked success against data from lived experience rather than compliance.

BBC Red Button Service provides on-demand international, national and regional news, sports reports, education and other information services via Digital Terrestrial Television (DTT)

In 1974 the BBC launched the world’s first Teletext based information service which evolved into the UK’s digital Red Button. Ceefax as it was called, had some great accessibility features for an analogue service, such as a high contrast interface, a form of stretched zoom mode, it was the basis of the first closed caption service in 1979 and there were adapters available for recording closed captions to VHS as open captions.

With the advent of DTT in the 1990s a new platform emerged which became the replacement for Ceefax, Red Button. This service is still available and has even got an enhanced version for internet connected TV’s and set-top-boxes, but for the purposes of this article I am focusing on DTT only version.

For roughly 5% of the UK’s population who cannot afford or use internet based services, this is still the only way they can access on-demand content. For many households broadband or mobile data are still luxuries and for others the whole connected experience is simply too complicated, and this is why the Red Button service is so important. Within this group we have some of the most disadvantaged and marginalised members of society for whom accessing the latest information on the COVID pandemic would have been very difficult without it. All households in the United Kingdom with a television have to pay a TV licence fee, and removing support for this platform would mean that it would create a two tier system for on-demand content.

People who have age related sight loss, reduced motor function or impaired cognitive function might not be able to learn a new and more complicated platform with operating system, browsers, websites and apps etc. They can simply be overwhelmingly complex, and using a system they have become familiar with over many years is key to providing equal access to information they have paid for when they want it.

The biggest issue the BBC had is that because Red Button on DTT is a push service with no return path, so finding and engaging users is tricky.

Because of this the users have no direct channel for their voices to be heard, and the BBC nearly closed the service in 2019 because it did not have the data at the time to ascertain the size of the audience and the role the service played in their lives.

When the BBC announced its intention, over 100 charities, large and small contacted the BBC to voice their concerns. They were able to provide the BBC with the data and insight it was lacking, and the service was saved.

One major learning that came from this is. If you can’t reach your whole audience, it might mean that conventional channels of communication are not available to them. It does not mean that they do not exist, but they are hidden in plain sight.

YouView is a hybrid television platform in the United Kingdom developed by YouView TV Ltd, a joint venture partnership of four broadcasters, the BBC, ITV, Channel 4 and Channel 5; and three telecommunications operators, Arqiva, BT Group and TalkTalk Group.

Back in 2009 I was approached and asked to write the accessibility requirements for this new TV platform. The project in question was called Canvas, which changed it’s name to YouView in 2010. It was a joint venture with the bulk of the design and engineering at the time being directed by the BBC but financed by the other partners. It was one of the most challenging briefs I have ever been given. YouView was not a standard set-top-box or a PC, and at it’s core it would offer access to free content delivered by IP. I was asked to make it as accessible as possible, but because this had to be affordable to lower income households, I was not allowed to add a penny to the price of the end product.

So I had to learn about what was intended, ask questions about technology capabilities and write requirements that were to be mandated by the BBC Trust that were also deliverable.

This is one of those moments where accessibility being built in from the start offered up a huge opportunity for innovation.

In the end we took a hybrid approach. At roughly the same time as the specification for YouView was being developed, the BBC was running the digital switchover Help Scheme which was part of the shut-down of national analogue broadcasting services. This was to provide and install for free new set-top-boxes for low income households, the elderly and disabled people, so everyone had access to the new services.

These were very basic DTT boxes but like with YouView there had been a brief to make them as accessible as possible and the lead engineer for accessibility Nick Tanton had some interesting ideas. We ended up working together to try and create similarity between the platforms, so people did not have to learn one from scratch if they had experience of the other.

At roughly the same time four commercial organisations were bringing out set-top-boxes or televisions that included spoken electronic programme guides. Panasonic, Samsung, TVonics and Harvard Electronics (Goodmans brand). At the time the licences for speech synthesis voices were incredibly expensive and the Goodmans and TVonics boxes had to be sold at such a premium to try and cover the cost of the licences hardly any sold any in the commercial marketplace. The Help Scheme stepped in to make them available as a subsidised upgrade, but even then the uptake was small.

Panasonic and Samsung, being much larger organisations were able to invest more into their projects and take a longer-term view of the feature, but they also struggled with another problem YouView had also identified and that was the difference between voicing an electronic programme guide, and providing voicing for third party applications for VOD services.

The problem was simply too complex and expensive for a smaller company like YouView and impact on the cost to all users would not be avoidable, without any guarantee of success.

The control would also be through the remote control. Remotes are not easy to make completely VI accessible because they are a shared not a personal device. As one blind research participant said, “if someone moves the remote control I am unlikely to be able to find it, and please do not make it whistle when I clap my hands because my house sounds like a rainforest of lost objects.”

The solution instead was to provide an API for all EPG and recording information so the box could be controlled and recordings could be set and played back using a dedicated app for iOS and Android devices. This would mean the device was less likely to be lost, the voice and navigation experience would continually improve with updates in the assistive technologies built in to their devices, where if it was all in the YouView box it would have been unlikely to be able to evolve in the same way.

Equally as there was no technical way of making text to speech viable directly on the YouView platform, but it was available on the upgrade boxes available through the Help Scheme, anyone who wanted a spoken EPG was better off buying one of their boxes, which was also a lot cheaper and so spoken access was available to lower income households.

This still left the issue of the platform apps such as BBC iPlayer or 4OD, and we decided at the time that instead of trying to fix the problem we would create an API standard that would enable any device to be able to synchronise and control any other device. This became known as the Universal Control API.

Some time great ideas just don’t work out, and this was one of mine. It was developed by some incredible research engineers at BBC R&D, and we continually tried but simply couldn’t get traction with the idea of a localised and universal accessible internet of things, and some of the opportunities that could have been realised, such as accessible second screen access to TV graphics, or a bbc.co.uk homepage that can update in sync with whatever you are watching, bit the dust, even though the prototypes were really promising.

We still were able to achieve a lot, but this all became about pushing the capabilities of the available technology to it’s limits.

I’ll start with what we were able to do to align the accessibility between the Help Scheme and YouView platforms.

We started with the obvious thing of supporting TV access services, namely closed captions and audio description, and providing dedicated buttons for each in the same place on both remote controls. As we couldn’t provide Text To Speech as standard across both platforms, we instead implemented an identical approach to audio feedback that used a series of beeps to let you know whether AD was available or not on a programme, and what the status of AD was on the box. All the beeps were at 1khz because using different pitches would have created problems for anyone who has both a severe vision impairment and congenital amusia AKA tone deafness. This is a condition that regularly gets forgotten in accessible audio feedback design despite it impacting 4% of the population.

On YouView we were able to then start focusing on many other groups of users, particularly those with impaired vision, sever motor conditions, cultural deafness and cognitive conditions.

Cognition

By now I had already become very interested in font accessibility, and had previously been involved in the BBC’s choice for an online font between Ariel and Verdana, where we had thankfully gone for Verdana, but I had learned a lot more since and wanted a font that would work for people with learning difficulties, dyslexia and vision impairment. We ended up settling on FS Me from Font Smith, a font that had been designed for MENCAP as their brand font. Their approach wasn’t particularly scientific, but the design process had been very inclusive. The font itself also sat well with the YouView brand and therefore became the interface font.

With respect to the core interface design we also considered other factors and balanced the needs between cognitive accessibility and vision impairment.

People who have irlen syndrome and even people with ADHD can be impacted by the presence of bright white or intense colour. The use of blue was part of a consideration to not visually overwhelm users, and there was full use of the pop out effect which can improve access for people with ADHD, especially with the main focal point being very obvious, related information that gives context to what is in focus is maintained in high contract and unrelated information falls back. When the point of focus changes so does the prominence of any related information.

To make the interface even more ADHD friendly you can turn off the live broadcast picture and turn any semi-opaque veils into solid ones so there are less distractions.

Cultural Deafness

For users who are Culturally Deaf we introduced a setting that means if a sign interpreted version of any programme is available it will play that version by default, as we had already learned in BBC iPlayer that finding sign-interpreted content can be tricky, because of the delay times introduced because of the time it takes to produce a quality interpretation.

Motor Impairment

One of the biggest issues from a motor impairment perspective is the remote control itself. As I have already mentioned, we had in some ways paired-up with the Help Scheme and followed many of the same best practices. We considered how easy the device was to grip, the contrast and clarity of the labelling, the size and distance between the buttons, having some of the core accessibility functions such as Zoom, Audio Description and Subtitles (closed captions) having dedicated buttons and nibs being added to other commonly used buttons, but this is only about optimising a device suing Universal Design that can’t avoid being full compromises. So we needed alternatives that gave comparative experiences.

The mobile application is one for people who had either an Apple or Android phone and knew how to take advantage of their accessibility features, but at that time those types of devices weren’t as prolific as they are now and they didn’t have the advanced switch control functionality they now have.

One of the requirements was to ensure the interface could be controlled with a single switch, and so we designed a Grid 2 based switch interface. We worked with users to identify their needs, one of which was for it to look as cool as the main interface. This could be controlled from a users existing set-up that acts as a remote control, giving users more autonomy.

Lastly we identified a middle ground of users whose needs were neither met by the remote-control or the switch software. On PC these users would often be keyboard navigators or large key keyboard users, so we decided that because USM keyboards were fairly cheap, we would utilise them as alternative input devices. We would map the functionality to mainstream keyboards as well as Jumbo and Clevy devices, and provided free downloadable overlays and sticker designs.

Vision Impairment

The final group we considered was people with vision impairment. The majority of this group were not dependent on speech output and so we thought about optimisation through a combination of six features.

1. The design of the main remote control being as optimal as possible with dedicated buttons for audio description and the zoom function.

2. The legibility of the chosen font. This is really important because it optimises some of the other features.

3. An audio feedback function that let users know if their button press has been received, and the status of things like Audio Description.

4. Because of how the core interface was being engineered we realised we could bring back an accessibility feature from Ceefax, a zoomed view. This evolve into ‘Interactive Zoom’, that enables both zoomed exploration but also interaction. A core feature is a dedicated button on the remote control. This is something that so many TV platforms fail at, is surfacing accessibility features in accessible places and too often bury them in inaccessible sub menus, which defeats the object of being accessible. Take a leaf out the Apple, Google and Microsoft product design books and make the path to accessible options as accessible as possible in the device’s default state, otherwise you are giving users what they need with one hand, whilst taking it away with the other.

5. A very simple feature was also added based on feedback from the RNIB which swaps out channel brand logos for labels.

6. And the last feature we also worked with the RNIB on and that was the provision of a high visibility alternative colour scheme. The main interface is high contrast for all primary and secondary information, but in this version it moved to high contrast for all text with some differentiation using background tones.

The point I’m trying to make with all of this is that to be effective in truly represent the whole demographic, accessibility has to be is a balance of pragmatism and idealism, married with an understanding of context, opportunity and audience impact. Accessibility has to think about both the audiences we know about and those that are hardest to reach.

Accessibility is often determined by a Western perspective of modern and expensive assistive technology that affords opportunity. But when you focus on the more profound challenges faced by disabled people outside of the top 30% richest people in the world, who have additional hardships, it is easier to see where the real digital divide is in terms of access to information, as well as learning opportunity and participation in society.

Additional Information

YouView Accessibility Pages

BBC Accessibility Pages and Resources

The Universal Control API has since fed into the spec for DVB-CSS https://dvb.org/?standard=dvb-companion-screens-and-streams-dvb-css-part-1-concepts-roles-and-overall-architecture

http://www.bbc.co.uk/opensource/projects/dvbcss-synctiming

https://github.com/bbc/dvbcss-protocols

#a11y #accessibility #diversity #uxresearch #uxdesign #inclusion #youview #bbc #worldservice