Automation is Not Equal Access
Melanie Thornton, Coordinator of Access and Equity Access
University of Arkansas – Partners for Inclusive Communities
If your goal is to provide equal access, it would be helpful to make this one of your mantras:
Automation does not equal access.
Whether we are talking about automated transcripts, automated conversion of documents, or automated testing of websites for accessibility, these processes cannot be relied upon to achieve equal access. Can automation assist you in your efforts to create accessible materials? Absolutely! But if you want to provide accessible digital materials, human intervention is still needed.
Automation and Video Captioning
Automatic Speech Recognition (ASR) has made great strides in the last 15 years. With the right conditions, people using ASR for dictation can achieve very high levels of accuracy. But the reality is that even if ASR were 100% accurate (and it is far from that!), it would not provide equal access to videos. A word for word translation of speech into text is not the same as captioning. It does not provide other sounds that are missing such as laughter or a knock at the door. It does not include any punctuation. It does not inform the reader who is speaking (if there is more than one person speaking). The reality is that accuracy rates vary widely but the average is about 65 to 75%. Eve Hill, a disability rights attorney, when asked if 65% is compliant, put it this way:
[Screen reader users: The following text has x’s scattered about to visually represent missing 35% of the text.]
“I XonXt kXow. IX onXy 7 oXt oX eveXy 10 XetXers Xere XorrXct Xn a bXok, wXuld Xou bXy XhXt bXok? XoulX thXt Xe XffeXtivX XommXuniXatiXn?”
It is really a disservice to call anything created with automation “captions.” Some people would even say that the automatic transcripts provided by some tech companies are a disserve to Deaf and hard of hearing people because many misunderstand and call it “done” without creating true captions. To create a truly accessible video, you can start with the automatically-generated transcript, but it is necessary to download it (or use online editing if provided), make corrections, add punctuation and other information that is missing, and then upload the caption file.
For more information on creating captions, see DMCP’s Caption Key and NCDAE’s tutorial on Captioning YouTube Videos.
Automation and Real Time Captioning
Since I began working in higher education almost 20 years ago, the question is often posed: “Can we just give the professor a mic and set up dictation software to create access for Deaf students?” The answer is the same now as it was then: “Automation does not equal access.” When someone uses dictation software as a writing tool, they have trained the software to their speech patterns. They are typically watching each word they write and making corrections if necessary as they go. They are also adding in punctuation and capitalization along the way. A professor giving a lecture will obviously not communicate in such an intentional way. Accuracy falls off dramatically and there are no punctuation marks. True access to lectures is provided by ASL interpreters, skilled CART (Communication Access Real Time) writers or, if the student prefers, speech to text service providers using Typewell or CPrint.
For more information, check out the National Deaf Center’s resources on interpreting and speech-to-text services.
Automation and Alt Text in Social Media
Just recently, several social media platforms have begun using AI (artificial intelligence) to create alt text for images. Is it better than nothing? It depends on the image. Most of the descriptions are not at all helpful. Memes with quotes or funny sayings on Facebook have automatic descriptions that read “Image may contact text.” There are occasionally more helpful descriptions such as “8 people smiling outdoors” but I did see a description of a pile of scrabble pieces that simply said “Image may contain food.” Clearly, sometimes the descriptions provided by AI are more confusing then if nothing was provided at all.
So again, we can’t rely on the automation to provide access. Currently, Twitter, Facebook, Instagram and LinkedIn all allow you to add descriptions to images that you upload. The links below will take you to instructions for adding alt text to the images in each platform.
While I could not find a tutorial for LinkedIn, they’ve made it pretty intuitive. When you upload a photo you are immediately prompted to add a description.
When you share images posted by others, there is not a way to insert alt text. To make those accessible, you have to add descriptions in the post or comments.
Automation and Document Conversion
Document conversion is another process that people often want to accomplish through automation. There are tools marketed that do automate the process of converting files from one format to another. While these tools can make the conversion process simpler, there are aspects of document accessibility that simply require human inspection and skills. Automation is no substitute. If the original document is not created in a way that is accessible, converting it will not change that. Documents that are complex and that have text boxes or images without alt text, will remain inaccessible if converted to another format. I would never recommend using an automatic tool and then handing the documents over to students who use screen reader or text-to-speech technology without close inspection and, if necessary, remediation.
Learn more about creating accessible documents.
Automation and Web Accessibility Testing
Not long ago, I shared some findings from testing a website for keyboard accessibility. A web designer informed me that such testing was unneccessary. According to him, all that was needed was to run the site through WAVE. Unfortunately, this is the dominant thinking about accessiblity. But even those who design automated tools agree that automated testing is only a part of the process of accessibility testing. Although estimates vary, most web accessibility experts agree that only 30% to 40% of the accessibility errors can be identified through automated testing. In order to ensure you website is accessible, a combination of processes are necessary including keyboard accessibility, automated tools, visual inspection and code inspection.
Learn more about accessibility testing.
Resource Highlight
Want to create a welcoming environment for people who use service animals? We provide complimentary fliers for you to download and post in your place of employment—college campus, shelter or any other organization.
We hope you find it helpful!
Upcoming Events
Free Webinar
Sponsored by Great Lakes ADA Center
The ADA, Businesses and Barrier Removal: What are the Requirements?
Tuesday, February 19, 2019
1:00 PM Central Time
Register for the ADA, Businesses and Barrier Removal Webinar >>
Free Webinar
Sponsored by Great Lakes ADA Center
Accessible Entrances, Doors and Gates
Thursday, March 7, 2019
1:30 PM Central Time
Register for the Accessible Entrances, Doors and Gates Webinar >>
Free Webinar
Sponsored by Great Lakes ADA Center
The Americans with Disabilities Act (ADA) and Web Sites: What is Required
Tuesday, March 19, 2019
1:00 PM Central Time
Register for the ADA and Websites Webinar >>
Free Webinar
Sponsored by Great Lakes ADA Center
What’s New with WCAG 2.1?
Thursday, March 21, 2019
1:00 PM Central Time
Register for the What’s New with WCAG Webinar >>
Stay Connected with Us!
Enjoy getting resources on access an inclusion? Follow Partners for Inclusive Communities on Facebook and Twitter for more news and resources!
About Explore Access
exploreaccess.org is a web resource developed through a project of the Southwest ADA Center Regional Affiliate – Arkansas, a program of UA Partners for Inclusive Communities through funding from the Southwest ADA Center (90DP0092).