You are here:

What did I learn today - CAPTCHAS

What Did I Learn Today

I am enrolled at Ryerson University, taking part-time distance courses towards a degree in Disability Studies.  I am registered with the Access Centre, Ryerson’s accommodation provider for students with disabilities.  This week, the Access Centre circulated a survey to students in an attempt to evaluate and improve the services they provide.   I thought to myself, here is an opportunity to recognize those things the Centre does well, make suggestions about those services where improvements should be made, and to discuss my views on accommodation versus accessibility.

I found all the radio buttons, filled in all the text boxes.  The survey was short, but we all know how difficult navigating a multiple choice survey can be.  I completed all the questions and pressed “submit”.  Nothing happened.  I pressed “submit” again, and again nothing happened.  Did I miss a question?  I moved backwards (I use a Mac so it is VO + left arrow - for those of you who use JAWS think of Shift + tab) until I found the message at the top of the survey - “incorrect CAPTCHA”. You can probably fill in the blanks for the expletive I uttered.

In order to submit my completed survey, I had to fill in the CAPTCHA.  I tried the audio CAPTCHA numerous times and each time shift tab to find the message that the CAPTCHA was incorrect.  So today, I learned about CAPTCHA.

CAPTCHA is an acronym for Completely Automated Public Turing test to tell Computers and Humans Apart.  CAPTCHAs are designed to allow humans to complete a web form, post a comment, or verify content.  CAPTCHAs are also designed to discourage automatic computer generated tools from accessing web materials, posting spam or entering some pages.  Unfortunately,  as stated in the W3C Working Draft 25 May 2011,  “all forms of CAPTCHA introduce unacceptable barriers to entry for users with disabilities” (

Audio CAPTCHA is not an accessible solution.  CAPTCHA audio tests are extremely difficult for a blind user as the screen reader speaks while the user is attempting to navigate through the controls, further distorting the already distorted audio output.  Some suggest that sighted users, without the additional interference of a screen reader, have less than 50% success using audio CAPTCHA (

In ‘Understanding WCAG 2.0’, W3C notes that CAPTCHAs are a ‘controversial topic in the accessibility community”.  Most organizations rely on CAPTCHA to limit the amount of unsolicited computer generated spam.  Organizations are understandably unwilling to give up the use of CAPTCHAs.  W3C  goes on to say that every type of CAPTCHA will be unusable by people with certain disabilities.  WCAG suggests that web pages should:

  • Provide more than two types of CAPTCHAs
  • Provide access to a human customer service representative who has the ability and knowledge required to bypass CAPTCHA
  • Not require CAPTCHA for authorized users

So I have written to the Access Centre and explained that they have sent a student registered with the Access Centre a survey about the Access Centre that is not accessible.  I will let you know how they “accommodate” this request.


This blog is curated by the AEBC, but welcomes contributions from members and non-members alike. The thoughts, views, and opinions expressed in the Blind Canadians Blog are those of the contributing authors and do not necessarily reflect those of the AEBC, its members, or any of its donors and partners.


How the B... hell do you fill in the blanks??

CAPTCHAs are something of a necessarily evil on the web thanks to all the spammers and garbage users out there. On the AEBC site, we employ text-based CAPTHCA questions (for example, 'Cow, bird, and paperclip: which of these is not an animal?'), which tends to serve our target population fairly well, though could present a barrier to some with learning or intellectual disabilities. CAPTCHA's generally are not employed for registered users (i.e. those who create and login with an account), although a visual/audio CAPTCHA may occasionally appear if the system thinks a comment on a blog post or other article is spam.

We do have the ability to bypass these tests altogether for individual users if the need arises as well, but so far it would seem most have not run into this difficulty.

Personally, I've never had a lot of luck with the audio CAPTCHA's. My hearing is supposedly normal but I have a very hard time deciphering those. I've never seen it as a particularly suitable alternatively to the visual CAPTCHA, especially since those two together do nothing to help those who are deaf-blind.

ZZ - Disregard this link; it is used to trick spammers.