Address-Based Sampling: Census Block Group Data Used to Define Incentive Structure
Anh Thu Burks
The Nielsen Company
Michael W. Link
The Nielsen Company
Within Address based sampling (ABS), limited focus has been on the vast array of demographic information available to researchers within this sampling frame. Appending demographic information to an ABS frame is a relatively simple process especially since it's comprised of address information for almost every sample unit. As examples, there are socio-economic variables, owner- renter status, household size, occupational and other Census-type variables that are available through sample vendors and can be used by researchers to tailor the survey recruitment process (i.e., mailings and incentives) in an attempt to gain respondent cooperation. Thus the focus here is to examine Census-type data at the block group level for each test case and how these types of variables were subsequently used to target incentives to specific demographic groups. Lastly, a test was conducted using an ABS sampling frame where participants were mailed a survey with a differential incentive and then provided the option to respond by web or mail. The findings from this test are discussed in terms of its effectiveness as indicators to target and gain cooperation among the hard-reach demographic groups (i.e.,18-34 year olds, blacks and Hispanics). In theory, this should help to compensate for the traditionally lower responses observed with these subgroups, serve to enhance the recruitment strategy and begin to address the issue of achieving better representation among these demographics. Findings here, suggest that indicators like Census-type variables provide a viable alternative to improving participation and to further improve sample representation among subgroups.
