RESEARCH TRIANGLE PARK, N.C.—The ability to collect household survey data using scientifically valid methodologies has long been a painstaking and expensive process for researchers seeking information on the opinions and behavior of Americans.
Now an innovation by researchers at RTI is being recognized for its ability to improve the efficiency and lower the cost of surveys by estimating and increasing the accuracy of household mailing lists.
The innovation was recognized by the American Association of Public Opinion Research with a 2011 Warren J. Mitofsky Innovators Award at the association's annual meeting in Phoenix, Ariz.
The award recognizes new theories, ideas, applications, methodologies or technologies that advance the field of public opinion research.
Prior to the innovation, those seeking to design accurate household surveys relied on on-site enumeration – physically visiting each household – to validate addresses. The time and expense associated with on-site enumeration often precluded it from being a viable option.
The primary alternative method, using residential mailing lists, had significant drawbacks because the completeness and accuracy of the lists was unknown.
To address this concern, RTI researchers evaluated the coverage of mailing addresses used as a sampling frame for a probability-based survey of 15,000 households in Dallas County, Texas.
Their results demonstrated that while there is under coverage (addresses of persons who request that their names be removed from a mass mailer's list), the use of statistical methods can estimate and increase the coverage of mailing lists.
As a result, researchers can increase the efficiency of the sampling design by identifying and possibly oversampling addresses adjacent to gaps in the delivery sequence.
A study outlining the innovation was published in Public Opinion Quarterly in 2003.