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

Web Survey Bibliography

Title Identifying and correcting question-wording problems: the case of Wageindicator
Year 2015
Access date 04.06.2015
Presentation

pdf (790 KB)

Abstract

This paper deals with identifying and correcting errors related to question wording in the questionnaire design step of the survey process. Each question can be worded in numerous ways which are not equivalent in terms of comprehensibility - some wordings are more familiar than others. In practice, absolute frequencies in language corpora can be used as estimates of wording familiarity. Supposedly, wordings with a higher frequency are more comprehensible to the respondent. Words with low frequencies can be replaced with synonyms or other similar words that can be found in thesauri. We took the Wageindicator Questionnaire as a case study to explore how to use these linguistic resources to identify and correct errors in question wording. 
The WageIndicator survey is used to collect data on wages and labour conditions in different countries. It is an online survey that uses a convenience, self-selected non-probability samples. The word count of the English version of the questionnaire is 6.181 words. We checked the word frequency of all nouns, verbs, adjectives and adverbs and made a list with all those that had a word frequency (based on the British National Corpus) in the lower 25%. We looked up for synonyms and other possible alternative words for these words and made a final selection of wordings that we used in a wording experiment: the original wordings were submitted to the control group, while questions with changed wordings were asked to the experimental group. We measured response times, drop-outs, item non-response, satisficing and respondent satisfaction in both groups. Based on the preliminary results we recommend further research on the potential of language resources that could help survey practitioners to identify problematic question wordings and suggest corrections.

Year of publication2015
Bibliographic typeConferences, workshops, tutorials, presentations
Print

Web survey bibliography - Slovenia (48)