Do multiple contacts between waves increase response, by reducing the proportion of untraced movers and increasing the sense of ‘belonging’ the sample member has with the study? Or could it have a negative effect, by annoying sample members and/or giving more chances to refuse? At the time of this experiment, Understanding Society had recently moved from one between-wave mailing per year to 3-4 mailings. This experiment attempted to gauge the effect this had on response at IP7 by allocating households at random to receiving one vs 2-3 mailings between IP6 and IP7. The standard procedure at the time was for the IP sample to be sent one between-wave mailing each year – usually around November. This experiment proposed that one half of households were sent three mailings in this period (September, November, February). The mailing content was the same that was produced for the main-stage mailings. This required that the random allocation for the mailings occurred earlier in the process for IP7 than scheduled (usually early-December), and had to be based on the IP6 sample rather than the fed-forward IP7 sample. Households within PSUs were randomly allocated to a control group and a frequent mailing treatment group.
Controlling variable(s) is ff_mailingsw7 on record hhsamp:
1 Control Group
2 Frequent mailings
