Notice: the WebSM website has not been updated since the beginning of 2018.

Web Survey Bibliography

Title Use and non-use of clarification features in web surveys
Year 2003
Access date 10.05.2004
Full text

pdf (364k)


Survey respondents sometimes interpret the words in survey questions differently than intended by question authors and this can lead to inaccurate answers. One way to clarify the intended meaning is to make definitions available. In Web surveys, it is possible to link words and their definitions so that respondents can click for clarification if they realize they might benefit from doing so and are willing to perform the necessary steps. In this paper we report an experiment in which we varied the familiarity of question terms (e.g. antioxidant vs. beer), the informativeness of the definitions (e.g. vegetables includes French fries vs. beef includes meat from cows) and the number of clicks required to obtain definitions (one, two or more). The experiment was embedded in a survey administered to a sample obtained from Survey Sampling Inc., yielding over 2,500 respondents, randomly assigned to the different conditions. Respondents were asked four questions about their consumption of food and nutrition. Overall, respondents were unlikely to obtain definitions (only about one sixth of respondents obtained one or more definitions) suggesting that many misconceptions go uncorrected. When respondents obtained at least one definition, they did so far more often for unfamiliar than familiar terms, indicating that they rarely recognize mismatched interpretations of words that they believe they know. On those occasions when definitions were obtained for familiar terms, they were more often informative than uninformative suggesting that respondents recognized the mismatch; the pattern was reversed for unfamiliar terms. Finally definitions were more likely to be obtained the fewer the number of required clicks reflecting respondents? willingness to expend truly minimum effort to clarify meaning. We discuss the results in terms of designing interactive Web survey features that are likely to be used, thus reducing measurement error.

Access/Direct link

Homepage - conference (abstract)

Journal Homepage (abstract) / (full text)

Year of publication2005
Bibliographic typeJournal article
Conferences, workshops, tutorials, presentations

Web survey bibliography - Tourangeau, R. (45)