Saturday, November 12
Questionnaire Design
Sat, Nov 12, 9:00 AM - 10:25 AM
Orchid C
Experimental Manipulation of Scales

The Use of Labels on Showcards with Many Response Options: An Explorative Study Using ESS Data (303591)

*Salima Douhou, City University London 

Keywords: questionnaire design, showcard, labels, face-to-face survey, cross-national

The use of show cards is common for closed-end items in face-to-face surveys. The European Social Survey includes an item on educational attainment and the Showcard for this item contains a list with many items and long labels that respondents might be less likely to read out completely during the interview. Several countries within the ESS use labels while others do not use labels on the Showcard of the educational attainment item. There are also countries that switched from no labels to using labels between rounds of the ESS. From a cross-national research perspective, it would be very relevant to find out what we should recommend: labels or no labels on these type of show cards.

We hypothesize that the cognitively challenging task of choosing a response option from a list with many items and long labels can be simplified by labelling the items. In addition, recency or primacy effects are expected in selecting from long lists.

We will report on two experiments: 1) An experiment in the pre-testing phase of the ESS in 2015 in two countries (UK and Estonia). The sample size is about 400 cases and about half of these respondents were randomly assigned to the group with a labeled Showcard for educational attainment and the other half received the Showcard without labels. 2) Natural experiments: Some countries in the ESS switched from no labels to using labels in the Showcard (each having sample size of about 1,500 respondents or more) as a result of organizational changes.

We report on DK responses, how scales are used by respondents in each treatment (increased use of top, bottom or middle categories?), item nonresponse and look at response times.