In 2006, U.S. Census Bureau staff conducted two iterations of field testing of a survey instrument that, in part, examines household mobility and identifies people with multiple residences. In July, staff field tested a computer-assisted personal interview (CAPI) version of the survey instrument in selected areas around Austin, Texas and the Cheyenne River Sioux reservation in South Dakota. In November, many of the same questions were asked in a small-scale nationally-representative (excluding Alaska and Hawaii) computer-assisted telephone interview (CATI) random-digit-dial (RDD) survey. Both surveys collected data using basically the same set of questions to gather a household roster and to capture other addresses where people in that household were reported to have stayed during that year.
This paper compares the demographic and address data collected in the two studies, comparing results from the RDD study to results from the site test. When we examine the trends in the demographic distributions of the two studies, we highlight how the findings follow the patterns expected for RDD and site test data based on the literature. To our knowledge, comparative literature between RDD and site test data for the main variables of interest in our studies (i.e., household mobility and the identification of people with multiple addresses) does not exist. This paper provides a starting place to look at the patterns of mobility and multiple addresses for these two field pretest methods. The results suggest that using a RDD method to study mobility and multiple residences might be comparable to using a site test.
RDD, CATI, CAPI, field test, mode effects, residence status, within household coverage
Nichols, Elizabeth, Jennifer Hunter Childs, and Kyra Linse. (2010). RDD versus Site Test: Mode Effects on Gathering a Household Roster and Alternate Addresses. Statistical Research Division Study Series (Survey Methodology #2010-06). U.S. Census Bureau. Available online at <http://www.census.gov/srd/papers/pdf/ssm2010-06.pdf>.