An ongoing concern over survey quality with focus on online sampling and survey design demands a broad analysis of all components of the process. To identify what matters and to what degree it impacts survey data quality, it is necessary to take all the individual components of this broad perspective and break them down into specific parts taking special care to look within each area for potential quality drivers.

We know online sample can be managed improperly and lead to biased results. We know survey design can be done improperly and skew responses. We know the interview length with online surveys can lead to satisficing or incomplete response, higher drop-out rates, and less satisfaction with the online survey respondent experience. We know there are many components to providing quality and we strive to do 51 what we can as an industry to effectively determine adequate procedural guidelines as we forge ahead with online survey research. For the purpose of this research presentation proposal we have chosen to present findings related to respondent engagement with particular attention paid to the role of grids in the research design and its impact on quality.

Our research findings are from a large internal R&D project conducted with our proprietary online research panel, Viewpoint Forum. In this research we conduct a multi-cell test to identify alternate ways of presenting attributes and the effect of the alternatives on data quality and respondent engagement. Key findings suggest a specific approach to survey design to maximize both data quality and engagement.

This presentation is focused specifically on this grid research and we hope to contribute to the larger body of knowledge that exists for how to effectively design a research study for an online modality.