Coverage bias in variances, associations, and total error from exclusion of the cell phone-only population in the United States

Abstract

Although landline telephone household surveys often draw inference about the general population, a proportion with only cell phones is excluded. In the United States, like in much of the world, this proportion is substantial and increasing, providing potential for coverage bias. Studies have looked at bias in means, but undercoverage can affect other essential statistics. The precision of point estimates can be biased, leading to erroneous conclusions. Research examining multivariate relationships will be further affected by bias in associations. A national landline telephone survey was conducted, followed by a survey of adults with only cell phones. In addition to estimates of means and proportions, differences were found for variances and associations. Bias in some point estimates was reduced through poststratification but became larger and in opposite direction for others. Different uses of survey data can be affected by omitting the cell-only population, and reliance on postsurvey adjustments can be misleading.