In part one, began our three part series on user accessibility by discussing the Design and Construction stages of your site and attempted to convey the importance of a trusted partner who will be handling those critical foundational steps. In part two we talked about the actual content going on your site and hopefully gave you some tips on how to deliver your message in a way that all users can appreciate.
Today in the final installment, we’re shifting our focus to the last two pillars: Testing and Feedback Loop.
Comprehensive Testing Program
Testing Websites for Accessibility
Testing can be grouped into two main categories: automated testing and manual testing. Automated testing can be defined as any testing that is done programmatically with software, whereas manual testing is done physically by a human. Both methodologies have their place and should be utilized when attempting to create or maintain an accessible website. Below we will touch on some of the unique strengths and weaknesses of each method.
Using an automated testing tool is often the first step taken to identify accessibility issues with a website and for good reason. Automated tools have the advantage of being able to check large amounts of content in a short amount of time. Some automated testing tools operate as a scanner or web crawler allowing them to traverse entire websites as long as there is a link to follow to the asset or page. Other tools take the shape of plugins that can only check a single page of a site at a time, or frameworks and APIs that require an advanced technical skillset to implement.
When choosing an automated tool possibly the most important thing to consider is not the breadth of items it can check for or number of guidelines covered, but rather the quality of the reports the tool is able to generate. The reports are going to be your guide to locating and addressing any accessibility issues so it’s paramount that they are descriptive. A good report will contain a summary of any issue found, where precisely the issue lies (page and line number(s), and what guidelines were used to determine there is an issue. Accuracy is also obviously key.
As mentioned above one of the main benefits of automated testing is the fact that automated scanning tools are able to check entire websites in a short amount of time. If you are working with a larger website it almost becomes a necessity to use some type of automated scanning tool. Many more basic issues can also be tested with great accuracy, such as checking images for the inclusion of ALT text or determining that forms are built correctly.
A problem arises where an automated checker may programmatically determine that something is correct. Software cannot determine things were done in a thoughtful and accurate manner. While ALT text maybe present on an image, and therefore pass an automated scan, the ALT text may not be accurate or descriptive enough to be helpful to users with disabilities. This one example brings us to the importance of manual testing.
Manual testing is any testing that is conducted by a human to identify accessibility issues. Inherently this type of testing will require a good deal of web development technical knowledge as well as a knowledge of accessibility issues and guidelines. A good way to conduct manual testing is to attempt to utilize your website the same way a disabled individual might. As was mentioned in the Knowledgeable Design and Site Construction section, a simple test is to see just how much of your website you can navigate to by only using the tab, enter, and arrow keys on your keyboard. A website that is highly accessible should have no problems being navigated only using the keyboard.
A manual tester will also utilize a screen reader to navigate and identify issues with a website. A tutorial about using the popular free screen reader NVDA can be found here: http://webaim.org/articles/nvda/. A manual tester will not only be able to identify violations of accessibility guidelines, but also provide real feedback on the usability of a website which is something an automated scanner will not be able to do. Ultimately manual testing is something that should be conducted by individuals who are experienced with the subject in order to ensure accuracy of findings. In other words, is the message being conveyed the intended message, or simply a message that conforms to the guidelines.
Easy Feedback Loop
We believe that one critical component of your effort should be providing impacted users with a clear path for reporting issues that they find. At your very best, even when employing multiple testing methods, the risk will always exist that something is missed. There really is no substitute, both from a testing perspective as well as a goodwill perspective, to let your clients know that you are doing your best to serve them and if they see something you can improve on to let you know. Our design team has recently adopted as a best practice and incorporated into our processes, to recommend that our clients include a webpage that allows the financial institution to declare their intent to adhere to the standards and provide very clear contact information for members/customers to easily contact the FI if something needs to be addressed.
In closing, don’t get overwhelmed by this. Find a vendor you trust that exhibits a good understanding of the requirements, but you cannot simply rely on a vendor. Does it change your processes? In all likelihood, yes it does. It may require that some of your staff because a part of your manual testing team to periodically check your site. You will need some method of automated testing that can be done far more frequently than a manual testing program will allow. You, as a financial institution, need to begin to view your messages and how they are delivered, from an additional perspective. But all of this, in short order, can become second nature to you if you make a plan for it. In the words of Benjamin Franklin, “Failure to plan is planning to fail.”