2010 JSM Presentations
ON THE QUALITY OF ANCILLARY DATA AVAILABLE FOR ADDRESS-BASED SAMPLING
Charles A. DiSogra, Knowledge Networks; Email
J. Michael Dennis, Knowledge Networks; Email
Mansour Fahimi, VP Statistical Research Services, Genesys Sampling Systems
A feature of address-based sampling (ABS) is versatility of the sample frame where many ancillary data can be appended to an address. Commercial databases, e.g., Experian, infoUSA, Acxiom are used to append observed and modeled information at various levels of aggregation. This enables researchers to develop more efficient sample designs and broaden analytical possibilities with expanded sets of covariates. While quality of ancillary data is of concern for researchers, the literature provides only anecdotal assessments on accuracy. Relying on surveys and KnowledgePanel® recruitment samples that employ ABS, the authors present results of comparisons between an array of ancillary data and corresponding observed values collected directly from the responding households. The same ancillary data are also used to demonstrate the ability to analyze non-response bias by comparing the ancillary data available for the invited sample and the subset of recruited study participants.
RESULTS OF TARGETING PRE-IDENTIFIED MINORITY, UNIDENTIFIED NON-INTERNET AND VACANT HOMES IN TWO NATIONAL ABS SAMPLES
Charles A. DiSogra, Knowledge Networks; Email
Erlina Hendarwan, Knowledge Networks, Email
Address-based samples are used to recruit the probability-based KnowledgePanel®. Bilingual invitation mailings ask households to do weekly online surveys for an indefinite membership period. Non-Internet homes are provided laptops and Internet access for participation. In two large samples, targeting is used to for minority and non-Internet homes. Using census data, addresses are pre-identified as African American or Hispanic. A combined 23% increased yield resulted. Non-Internet homes cannot be pre-identified so a "target note" is included with a mailing to non-responders to reach these otherwise unidentified homes. Yield of non-Internet homes doubled, 11.4% to 22.1%. To maximize coverage, addresses classified as vacant were added to both samples. The vacant are examined for current status based on undeliverable mail and found recruits. Yield is one-fifth that of non-vacant.
