A dilemma is any situation where the decision to be made between two outcomes results in an inevitable cost. A dilemma is the choice between the better of two evils. The down sides of a web development dilemma (often referred to as a trade-off or compromise) may include the cost of implementation and maintenance, the inclusion of one group to the exclusion of another group, or impacts on website performance and effectiveness. Our skillset is the mastery of intelligent compromise.
Website accessibility is fundamentally an ethical responsibility shared by the business owner and the web developer. Individual strategies we implement to assist one group of disadvantaged customers can expose us to ethical dilemmas by creating barriers to other groups who are equally deserving of access.
The classic example of an accessibility solution creating accessibility barriers was the WCAG 1.0 requirement that developers provide keyboard shortcuts or accesskeys. This was not included in WCAG 2 in 2008 because real world experience revealed that accesskeys over-rode default screen reader short cuts. To not implement accesskeys had a down side; and to implement accesskeys had an equally powerful down side. This fit the criteria for a dilemma. The lesser of two evils was to “do no harm.”
A closely related problem that we have to deal with is accessibility complacency. The temptation to use WCAG 2 as a tick box to claim a website is accessible. That binary idea of accessibility is as dangerous as not realising the underlying issue. Accessibility is aspirational and we need to continually reinvent the way we look at problems to find the best solutions, particularly in cases that could be defined as ethical dilemmas.
A relevant contemporary example of this struggle for compromise between the lesser of two evils can be seen in user level verification when logging into online banking systems. Solutions often include the customer being asked for the 3rd, 7th and 9th digits in their account number, or expected to remember and select 3 of 9 available icons in a specific order from memory.
The first solution will heavily impact 3-6 per cent of our population with dyscalculia (a roughly equal sized group to dyslexia). Dyscalculics experience an inability to cope with abstract mathematical concepts we generally take for granted like time and simple addition or subtraction. Dyscalculia is a physical brain issue and carries with it legal implications that need to be appreciated. This first solution also poses significant challenges to anybody with cognitive disability. However, this solution does service most people, including those who can access the numbers as text using assistive technologies.
A web manager being asked to implement user level verification should see a down side to using this solution. But in reality, accessibility decisions are a compromise with some groups being helped into the banking system and other groups being excluded as a result of the design.
The second solution, using icons, is of equal concern because it requires visual cues and customer memory. Customers with poor vision are being asked to access visual icons, while anybody with cognitive disability may not be able to remember one icon from the next. This is without considering the size of icons, their affordance and the ability of some groups to physically click the interface. Colour blindness could confuse between icons, or cultural conflict between the meaning of both colour and the icon design.
Regardless, both solutions ring uncannily of the CAPTCHA.
At this point, the web manager has two bad decisions and is being asked to choose – an ethical dilemma. Does the web manager go to Kant for a universalisable maxim in black and white absolutes? Or to Utilitarian Analysis choosing the greatest good for the greatest number? Rights Analysis would suggest the ethical considerations are clashes between positive and negative rights. And the Justice approach would ask the least advantaged party what they would decide? This is the meat of an ethical dilemma.
No right answer exists and various ethical theories suggest different approaches to the problem. Do we implement solution A and accept the down side, or solution B and accept the down side? Which group of customers (dyscalculics or vision impaired, for example) will be facilitated and which group ostracised within the system we develop?
So I offer a solution. I don’t offer it lightly. It will come at expense and it will consume your time. Your client may balk, but you have an ethical responsibility to consider it. That solution is a “cascade of inclusiveness.” It is the idea that we consider a multi-levelled approach to web accessibility that tries to identify and overcome ethical dilemmas where they appear. This question of solution A and B is contrary to our underlying ethical responsibility to the broader community.
In this banking scenario the web manager could turn to the business manager and refuse to accept the limitations of the two-choice dilemma. Solutions outside the square fit into this category; so does offering multiple options to customers to broaden the individual gateway into the banking system.
In this case, the customer can be asked to choose either A or B or any number of contrived equivalences that meet our minimum criteria. The customer determines, to the best of their personal abilities, how they wish to pass any secondary identification test. After all, the burden isn’t the customer’s to bear, but ours. The solution needs to be in the design of things, wherever possible.
Yes, the cascade of inclusiveness is entirely obvious. And yes, it may be considered redundant and expensive by decision makers.
But call me an optimist. The cascade of inclusiveness will improve conversion rates, exceed legal requirements and has marketing potential. That is a compelling argument in a business meeting.
I shoot film photography for fun. I have an MBA (Specialisation) and a Bachelor of Computing from the University of Tasmania. Currently in pursuit of investment for a local business venture. Dreams of owning the World. Idea champion. Paradox.