[lbo-talk] Pew study of poll accuracy

Doug Henwood dhenwood at panix.com
Tue Apr 20 13:15:57 PDT 2004


Polls Face Growing Resistance, But Still Representative

Survey Experiment Shows

Public opinion polls have become harder to conduct than just a few years ago. Yet a new survey experiment shows that carefully designed and implemented polls continue to obtain representative samples of the public and provide accurate data about the opinions and experiences of Americans.

The study by the Pew Research Center found that the typical five-day survey, using standard techniques, now reaches just 27% of sampled households. In a similar experiment in 1997, a comparably designed standard survey completed interviews with 36% of sampled households.

This decline in participation, while notable, has not undermined the validity of surveys conducted by national polling organizations. When compared with benchmarks obtained from the U.S. Census and other government surveys with extremely high response rates, the demographic and social composition of the samples in the average poll today is highly representative. Moreover, there is little to suggest that survey respondents and people who do not participate in surveys hold substantially different views on political and social issues.

Surprisingly, the Center has found that the growing use of answering machines, caller ID and other call-blocking tools is not preventing survey organizations from reaching an adult in most of the households sampled. But because of outright refusals or 'soft' deferrals, interviews were completed in just 38% of the households reached with standard polling techniques -- down from 58% in 1997.

As in its 1997 study, the Center has found little difference between a standard survey -- conducted with commonly utilized polling techniques over a five-day period -- and a survey conducted over a period of several months employing much more rigorous techniques. In fact, there were no major attitudinal differences between participants in the standard survey and even the hardest-to-reach respondents -- those who were reached only after multiple attempts or had declined to participate on at least two occasions.

View Report: <http://people-press.org/reports/display.php3?ReportID=211>



More information about the lbo-talk mailing list