Notice: the WebSM website has not been updated since the beginning of 2018.

Web Survey Bibliography

Title Coverage Bias in Surveys Excluding Cell Phone Only Adults: Evaluation of Bias and Effectiveness of Post-Survey Adjustments
Source The American Association for (AAPOR) 63rd Annual Conference, 2008 & WAPOR 61th Annual Conference, 2008
Year 2008
Access date 22.05.2009
Abstract

While landline telephone household surveys often draw inference about the U.S. population, a proportion of adults with only cell phones are excluded. This proportion is substantial and increasing, providing potential for coverage bias. There is a need to improve understanding of coverage bias and the ability to adjust for it. Studies have looked at bias in means and proportions, but undercoverage can affect other essential statistics and uses of survey data. Even in the absence of bias in means, their precision and group comparisons will be misestimated when variances are biased. Much of research focuses on multivariate relationships in the population, which can be affected by bias in associations. The expected direction of bias can be informed by other research, through characteristics like younger age. This can be misleading. Counter to expectations, coverage bias may even be in the opposite direction when controlling for these characteristics. Finally, coverage bias is suspected as the cell-only population is different on demographic characteristics. These characteristics are commonly related to survey measures, creating the necessary conditions for bias. Yet bias is probable after adjustment only when differences on survey variables exist within demographic groups. A national landline telephone survey was augmented with a cell phone sample. Differences between samples were found in estimates of means, proportions, variances, and associations. Bias in some means and proportions was reduced through poststratification, but became larger and in opposite direction for others; cell phone respondents were more likely to report victimization, but conditional on demographic characteristics, were less likely to report it. We conclude with implications, cautions and suggestions for future work.

Access/Direct link

Homepage (abstract)

Year of publication2008
Bibliographic typeConferences, workshops, tutorials, presentations
Full text availabilityAvailable on request
Print