Accessibility

As a general rule, SoSci Survey creates accessible questionnaires. However, accessibility requires technology and content to work together – therefore, we have put together some recommendations here.

To begin with: Accessibility is just as multifaceted as the spectrum of possible individual impairments. People with physical impairments often rely on assistive technologies (for example, screen readers, alternative input devices). The questionnaire does not have to provide much additional support here: Often it is sufficient if it does not interfere with the assistive technologies doing their job.

The following notes focus on people with visual impairments and with cognitive impairments.

Important: Accessibility is not only useful to avoid systematic bias. In many cases, accessibility is required by law – for example, also for universities, and this may also apply to research that students conduct as members of the university.

Tip: Clearly worded, specific questions and response options are not only helpful for people with impairments. Short sentences and concise language help all respondents and improve the reliability of the survey. Ideally, remove irrelevant and repetitive texts from the questionnaire right from the start.

SoSci Survey creates HTML code that screen readers can handle well, and by default offers buttons for increased contrast and larger font size. In addition, the default layout of SoSci Survey uses the font Atkinson Hyperlegible, which the Braille Institute developed with the aim of particularly good legibility. Furthermore, users of screen readers can switch the questionnaire to a simplified display mode, which, for example, displays ranking questions with dropdowns instead of cards to be moved.

In experimental studies, pay attention to your stimuli: If these are images, inform people with visual impairments early on that it is not possible to complete the questionnaire with a visual impairment. Under Compose Questionnaire in the Settings, select the option “Questionnaire not suitable for screen readers due to method” for the Screen reader mode.

Do not use images if your stimulus actually is formatted text, for example tables or a curriculum vitae. In such cases, use HTML code for formatting. AI bots can be of excellent help in creating HTML code, if HTML is a challenge for you. Unlike images, HTML code is usually also easy to read on different devices.

For detailed information, please read the manual Surveys for Visually Impaired Participants.

For people who have difficulty reading and for functional illiterates, SoSci Survey by default offers the option of having the texts on the screen read aloud. This function is activated via the speaker icon at the bottom. The function relies on standard functions of the operating system, which is why the voice sounds different on Android, iOS, Windows, and MacOS.

Using easy-to-read language (“Leichte Sprache” in German) goes one step further. If you would like to offer a questionnaire in easy-to-read language, you have to translate the questionnaire into it. Thanks to AI, you can have at least a non-certified translation created largely automatically, for example via https://optimeil.de/leichte-sprache/ (for German texts). However, you will notice that easy-to-read language is considerably longer than ordinary language. Remove irrelevant content – ideally in the original language version as well.

In SoSci Survey, you can provide such a translation just like any other translation, via the function for multilingual surveys. Simply select “User defined language” as the additional language.