Relevance & Research Questions: More and more participants respond on different mobile devices to questionnaires. Research describes this phenomenon as "unintended mobile participation". The drop-out rate in those studies is higher in mobile studies and the response time is considerably longer making research more expensive. While looking for new approaches in the sphere of HTML5 a joint team of researchers compared the behaviour of mobile and desktop respondents when confronted with modern technology with graphical support compared to a standard HTML questionnaire.
Methods & Data: With the support of a professional panel provider and the market research app of a major german TV company the study integrated more than 800 complete interviews in Germany. The recruiting took plave over three major technical channels: PC Browser, Mobile Browser, an Mobile In-App questionnaire. A control group received classic question types while the experimental group received HTML5 supported questiontypes for the identical questionnaire.
The comparisons included:
- Single Choice (without vs. with graphical support)
- Drop-Down (List vs. Graphical Scaling)
- Sorting Task (Drop-Down vs. Graphical Sorting)
- Price-Sensitivity-Meter (Type-In vs. Graphical)
- Analytic Hierarchy Process (Classic vs. HTML 5)
All elements were designed to be functioning device agnostic on all online devices in the same mechanics.
Results: The HTML5 questionnaire design allows faster responses, lowering the burden of mobile responses resulting in less drop-out. Some of the new question types show deviant result patterns compared to the classic approach of asking. While some of the new question types seem to drive research into a direction of higher validity, other ideas did not meet the expectations of the team looking for ways to interview people on all possible channels.
Added Value: The team is willing to share their findings in the field of mobile browser and In-App research approaches for a world with less problematic questionnaires.