Recent findings suggest a significant drop in health satisfaction across waves of the Understanding Society survey. It is unclear if it is an actual change or artefact of the survey design. The presentation of response options also changed across waves, in terms of orientation of the response options, but research does not suggest such a change should be expected. The present experiment uses three different presentations of question and response options on the standard set of satisfaction question used in the self-completion component on the IP. One version had all satisfaction items presented on one page, with response options across the top and the items on the left in a grid format. The second version had each question presented on a separate page, with response options vertically aligned. The final version also had each question presented on a separate page, but response options were presented horizontally. The goal of the experiment is to identify if and how the presentation of response options in self-completion formats can impact estimates of satisfaction.
The allocation occurred at the household-level, with 1/3 of households randomly assigned to each of the version. The controlling variable is
ff_satisw9 in file i_hhsamp_ip
1 Grid format
2 One question per screen, response options vertically aligned.
3 One question per screen, response options horizontally aligned.
The variables used for this experiment are in record i_indresp:
sclfsat1, sclfsat2, sclfsat7, sclfsato
