Web Survey Bibliography

Title Open vs. closed questions in Web surveys
Year 2002
Access date 04.05.2004
Abstract Two quite different reasons for using open as opposed to closed questions can be distinguished. One is to discover the responses that individuals give spontaneously, the other is to avoid the bias that may result from suggesting responses to individuals. However, open questions have also their disadvantages in comparison to closed questions, such as need for extensive coding and larger item non-response. While this issue has already been well researched for traditional survey questionnaires, not much research has been devoted to it in recently used Web questionnaires. We therefore examine the differences between the two forms of survey questions by means of experiments within large-scale RIS 2001 Web survey. Three survey questions were asked in two forms in a split-ballot experiment: question on most frequently visited Slovenian and foreign Web sites, e-shops where respondents made some purchase, and the most important, critical problem the Internet is facing today. The results show that in all cases there were differences between question forms in univariate distributions, though no differences were found in the ranking of values. Closed questions in general yield higher percentages than open questions for the answers that are identical at both question forms. It seems that respondents restricted themselves with apparent ease to the alternatives offered on the closed form, whereas respondents on the open question produced a much more diverse set of answers. In addition, our results suggest that open questions produce more missing data than closed questions. Even more, there were more inadequate answers for open questions (e.g. respondents mention Slovenian Web pages on question on foreign Web pages or it is impossible to determine what the answer is about). This suggests that open questions should be more directive in wording (at least for Web surveys as self administered mode of data collection) as closed questions that are more specified with given response alternatives. We also found out that when respondents were asked about a more salient topic the differences between open and closed form were smaller. In this case both forms of questions were equivalent.
Access/Direct link Homepage - conference (abstract)
Year of publication2002
Bibliographic typeConferences, workshops, tutorials, presentations

Web survey bibliography - Vehovar, V. (139)