Doctoral Research Dissertation: An Experiment to Examine the Effect of Visual Grouping in Helping Respondents Navigate Skip Patterns

University Of Nebraska-Lincoln, Lincoln NE

Investigators

Abstract

Mail surveys are becoming more popular among survey researchers because they can achieve better coverage and response rates at a lower cost than can be achieved with telephone surveys. However, there are still challenges when using mail surveys. This research project will examine a common design challenge that has received only minimal research attention: how to get respondents to correctly navigate skip patterns without computer or interviewer assistance. The study will advance the visual design literature within survey methodology from its previous focus on individual items to a new focus on how visual design can be used across questions. As such, it has the potential to broaden our understanding of how visual design can be used to design good questionnaires, not just good questions. Researchers from multiple disciplines and across all survey sectors will be able to utilize findings from this research in their own surveys to minimize skip errors, thereby reducing missing data and increasing statistical power. As a Doctoral Dissertation Research Improvement award, support is provided to enable a promising student to establish a strong, independent research career. Failing to correctly navigate skip patterns in mail surveys without computer or interviewer assistance often results in missing data. Recently, survey methodology has incorporated knowledge from the vision sciences and cognitive psychology into understanding how visual design can be used strategically in questionnaire design. This study uses an experimental design to examine the effect of visual design features that create stronger grouping and subgrouping among items within skip patterns on skip errors in a mail survey. The survey will be administered to a sample of Nebraska residents drawn from the Computerized Delivery Sequence File. Specifically, this study will examine how grouping questions through enclosure or common region and indentation affects the rate of skip errors. There will be four forms sent out. The control form will have all of the questions located in the same continuous region (as established by a common shaded background) and all questions started at the left margin. The second form has the same continuous region but all follow-up items will be indented. The third form will have all questions start at the left margin, but the follow-up items will be in a separate background region from the initial feeder question. Finally, the fourth form will examine the combined effects of grouping through the use of both enclosure and indentation. Rates of skip errors for each objective will be examined across respondent characteristics (i.e., age, education, literacy) and question types (i.e., number of response options, number of follow-up questions).