How the L.A. Times/Kaiser Family Foundation poll was done

The Los Angeles Times/Kaiser Family Foundation survey of adults with employer-sponsored health insurance is based on interviews with a probability-based sample of 1,407 respondents between the ages of 18 and 64 who reported having health insurance from an employer or union (excluding those covered by a parent’s employer).

Interviews were administered online and by telephone from Sept. 25 through Oct. 9, 2018, in English and Spanish.

Teams from The Times and the the Kaiser Family Foundation worked together to develop the questionnaire and analyze the data, and both organizations contributed financing for the survey. Each organization is solely responsible for its content.

The NORC independent research organization at the University of Chicago conducted sampling, interviewing and tabulation for the survey using the AmeriSpeak Panel, a representative panel of adults 18 and over living in the United States. AmeriSpeak Panel members are recruited through probability-sampling methods using the NORC National Sample Frame, an address-based sampling frame. Panel members who do not have internet access complete surveys via telephone, and internet users complete surveys via the web (for this survey, 60 respondents completed via phone and 1,347 via web).

For this study, panelists with lower incomes were selected at disproportionately higher rates than those with higher incomes in order to allow for separate analysis of low-income people with employer-sponsored insurance.

The combined results have been weighted to adjust for the fact that not all survey respondents were selected with the same probability, to address the implications of sample design, and to account for systematic non-response along known population parameters. The first weighting stage addressed differences in probability of selection for the AmeriSpeak Panel and accounted for differential non-response to the AmeriSpeak screening interview. At this stage, an adjustment was also made to account for the oversampling of lower-income panelists and to account for differential non-response to the survey screening interview.

In the second weighting stage, the sample was adjusted to match known demographic distributions of the U.S. population ages 18-64 with employer-sponsored health insurance using the following parameters: age, age by sex, age by education, gender by education, education within non-Hispanic whites, race/ethnicity, census division, and household income. Demographic weighting parameters were based on the U.S. Census Bureau’s March 2018 annual Social and Economic Supplement of the Current Population Survey.

The margin of sampling error including the design effect for the full sample is 3 percentage points in either direction. All statistical tests of significance account for the effect of weighting. Numbers of respondents and margins of sampling error for key subgroups are shown in the table below. For results based on other subgroups, the margin of sampling error may be higher.

Sample sizes and margins of sampling error for other subgroups are available on request. Sampling error is only one of many potential sources of error in this or any other public opinion poll.

Kaiser Family Foundation public opinion and survey research and NORC at the University of Chicago are both charter members of the Transparency Initiative of the American Assn. for Public Opinion Research.