Effects of showcards on responses in face-to-face surveys

Summary

There is currently very little empirical evidence on how showcards are
used in surveys and what effect they have on response. Some face-to-face
surveys use these visual aids to reduce the cognitive demands on memory
by letting the respondent read response options instead of having to
remember them. However, showcards could also make the task of answering
survey questions more complex, as the respondent has to listen to the
interviewer and read the showcards (at the same time). In this paper we
use experimental data from the Understanding Society Innovation Panel,
which contrasted showcard and no-showcard versions of questions. We have
tested 54 items using Wave 2 Innovation Panel data and 14 items using
Wave 1 and 2 Innovation Panel data. Using Chi-squared and logit test
statistics we calculate whether response distributions are different
between showcard and no-showcard conditions and whether the effect of
showcards is different for respondents with different levels of
cognitive ability. Results so far suggest that showcards only rarely
affect responses. Further analyses will include examining whether the
use of showcards affects the response times for questions. We also plan
to replicate the analyses on an experiment of the European Social
Survey, and results will be presented at the conference.