Two experimental conditions exist for this experiment. First is the medium of obtaining consent; the second involves the wording to the consent request. Respondents were first assigned to either an “opt-in” or “opt-out” consent condition. In the opt-in condition, respondents were asked directly in the IP10 questionnaire about whether they consent to link their electoral registry data to their survey responses. The opt-out condition informed respondents of linkage during the inter-wave mailing, giving them a chance to send in a Freepost response asking to be excluded from the linkage.
For both the opt-in and opt-out conditions, there were two wording versions for the requests. The first used the BES 2015 wording, and the second adapted it to be more explanatory. Random allocation to treatments occurred at the household level with equal allocations of 50/50 to each condition, such that the end result is there is a 25/25/25/25 assigned to each of the possible combinations of the two conditions. The controlling variables for this experiment are
ff_eleclinkinw10 in record j_hhsamp_ip:
1 Electoral register linkage opt-in experimental group
2 Electoral register linkage opt-out experimental group
ff_elecwordw10 in record j_hhsamp_ip:
1 Electoral register linkage wording experiment Version A
2 Electoral register linkage wording experiment Version B
Variables used for this experiment are in record j_indresp_ip:
eleclink_a, eleclink_b, elecadd, elecnewctry, elecoptout10
At Wave 11, respondents in the opt-out condition were asked about their memory of the interwave mailing between IP10 and IP11. These follow-up measures were asked only of those in the opt-out condition (50% of the IP10 sample). The measures asked about memory about the request in the interwave mailing, their belief whether they consented or not given the information, the usefulness of linkage, and acceptability of this method to ask for linkage. Finally these respondents were asked directly for their consent to link to their data in the electoral register, to ensure their desires are accurately expressed. Random allocation occurred at IP10 and is fed forward to IP11 (variable ff_eleclinkinw10 in record k_hhsamp_ip).
Variables used for this experiment are in record k_indresp_ip:
eleclinkmem, eleclinkcon, eleclinkuse, eleclinkacc, eleclink11
