Go to UPA home page The U P A voice
April 2007 Contents

UPA Job Bank

UPA 2007


Seeking an Accessible and Usable Survey Tool

By Karen Mardahl and Lisa Pappas

Karen Mardahl is a technical writer based in Denmark. She is co-manager of the AccessAbility SIG of STC along with Lisa Pappas.

Lisa Pappas is an accessibility analyst with SAS in Cary, NC. She authored the Institute's accessibility white paper: Universal Design: A Commitment to Accessible Software. She works with software testers to identify and prioritize accessibility issues, with development teams to address issues found, and with sales to complete accessibility checklists used in procurement.

When we set out to survey members of the AccessAbility SIG of Society for Technical Communication (STC), we needed an accessible tool to live up to the SIG's name and charter. Free was also a nice price tag.

Survey Monkey is the most frequently mentioned free survey tool among various special interest groups in STC, so our search started there. The very first test revealed a major accessibility stumbling block: it is not fully keyboard accessible. Keyboard accessibility is crucial, as assistive technologies such as screen readers and voice-input systems depend upon it. When we asked the support team at Survey Monkey about this issue, they replied that Section 508 compliancy was under consideration, but there was no known date for delivery.

Thanks to a tip from the STC Suncoast chapter president, Survey Gizmo was next on our list. It was definitely closer to accessible: keyboard navigation and submission was possible, it survived Freedom Scientific's JAWS screen reader nicely, and you can choose high contrast colors. High contrast support is very important for low-vision users. One drawback is that you cannot put alt text on a logo image (to promote your organization); however, you can omit that without losing value.

Tool Comparison

AspectSurvey MonkeySurvey Gizmo
Inheritance, Look & Feel (L&F) Choices Pros
  • Sufficient L&F choices to accommodate contrast needs
  • Inherits larger font setting
  • If user disables CSS, there's no change
  • Exit link lacks contrast
  • Readable / usable without a style sheet
  • Can create your own L&F template (so high contrast / large fonts)
  • Main page degrades gracefully when CSS is disabled
  • Access keys provided
Screen Reader
  • Graphic buttons lack alt text, not evident as controls when read aloud
  • Labels not explicitly connected to controls
  • Tables used for website layout confuse a screen reader
  • Labels identified except on follow-on text areas (may be user error)
  • Can step through form controls via keyboard;, all text read.
  • State of controls updates aurally when value changes
  • Visible focus not maintained.
  • No tab indices, so cannot navigate among controls to select values
  • Visual focus indicator maintained as you tab through controls.
  • Claim to be working on accessibility, but no time estimate for availability.
  • Free version limited to 10 questions / 100 responses
  • List competitors and compare pricing
  • Home page has screen-reader and keyboard friendly mechanism to skip over navigation links and access keys are provided.
  • Validator tool flags <br> tags, but that does not affect accessibility.
  • Where graphics are used to space layout (spacer.gif), the image tags need empty alt text values (alt=""), so that the screen reader will ignore them, rather than announcing the graphics as untitled.
  • Free version available 250 responses
  • Can import surveys from Survey Monkey

Because we found an accessible tool, we stopped our search and deployed a survey with Survey Gizmo. Survey creation and activation was simple, and the results were easy to gather. But most importantly for us, our users with disabilities reported no difficulties in accessing, completing, and submitting the survey.

Additional Comments

Lisa took the practical approach to validating as opposed to an automated check, which would probably flag accessibility issues with this survey. Automated tools can check code, but the results may provide false positives and will completely miss manual checks, such as whether alternative text for graphics is in fact meaningful.

Some users of assistive technologies disable JavaScript, so this is another factor to consider. Because support of dynamic web elements has improved in JAWS version 8, we decided to accept this risk.

Jim Thatcher's site has more information about making your forms accessible (accessed March 11, 2007).


Usability Professionals' Association
promoting usability concepts and techniques worldwide
Phone + 1.630.980.4997         office@upassoc.org

Contact the Voice