Web Survey Bibliography

Title Collecting Eye Tracking Data to Test QUAID, A Web Facility that Helps Survey Methodologists Identify Problems with the Comprehensibility of Questions
Year 2003
Access date 07.05.2004
Abstract When respondents do not understand the meaning of a survey question, it is very unlikely that they will supply valid and reliable answers. Survey methodologists would therefore benefit from computer tools and other analytical schemes that help them identify problems with questions with respect to comprehension difficulty. We developed a web facility (called QUAID) that assists survey methodologists in improving the wording, syntax, and semantics of questions (www.psyc.memphis.edu/quaid.html). QUAID stands for “Question Understanding Aid.” The survey methodologist enters a question on a questionnaire, along with any context information and answer alternatives that accompany the question. QUAID quickly returns a list of potential problems with the question, including: unfamiliar technical terms, vague or imprecise relative terms, vague or ambiguous noun-phrases, complex syntax, and working memory overload. Recent advances in cognitive science, computational linguistics and discourse processing have reached the point where it possible to analyze language and meaning at these various levels. Previous published studies have reported how well QUAID diagnoses the five categories of problems with questions in a corpus of questions provided by the US Census Bureau. Experts in language and cognition rated each question as to whether it had the five problems; the expert ratings were compared with QUAID output. Analyses of hit rates, false alarm rates, and discrimination scores confirmed there is a significant correspondence between QUAID and the judgments of experts. More recently, two eye tracking experiments were conducted that collected data from respondents as they answered questions. If a question is problematic, this should be reflected in the patterns of eye movements and the amount of time that respondents focus on individual words. The proposed presentation will describe the features of QUAID, the validity of the tool with respect to comparisons with experts, and the results of the eye tracking experiments.
Access/Direct link Homepage - conference (abstract)
Year of publication2003
Bibliographic typeConferences, workshops, tutorials, presentations
Print

Web Survey Bibliography - The American Association for Public Opinion Research (AAPOR) 58th Annual Conference, 2003 (45)