Methodology

The American Values Atlas (AVA) is a project of PRRI. All demographic, religious affiliation, and political affiliation data are based on approximately 50,000 bilingual telephone interviews conducted over a period of 52 weeks across a single calendar year. Beginning in 2014 and moving forward, the AVA will also include a set of issue questions comprising a subset of the total sample (roughly 40,000 interviews) covering a variety of topics.

Roughly 1,000 interviews are completed each week, with at least half the interviews conducted among respondents on their cell phones. In each week, interviewing occurs over a five-day period, from Wednesday through Sunday. The selection of respondents within households is accomplished by randomly requesting to speak with the youngest adult male or female currently living in the household. Interviewing is conducted by professional interviewers under the direction of SSRS.

Data collection for the AVA is based on stratified, single-stage, random-digit-dialing (RDD) sample of landline telephone households and randomly generated cell phone numbers. The sample is designed to represent the total U.S. adult population (age 18+) from all 50 states, including Hawaii and Alaska. The landline and cell phone samples are provided by Marketing Systems Group.

The weighting for AVA is accomplished in two separate stages. The first stage of weighting corrects for different probabilities of selection associated with the number of adults in each household and each respondent’s telephone usage patterns—telephone usage refers to whether respondents have only a landline telephone, only a cell phone or both types. In the second stage, sample demographics are balanced to match target population parameters for gender, age, education, race and Hispanic ethnicity, region (U.S. Census definitions), population density, and telephone usage. The population density parameter was derived from the latest available Census data. The telephone usage parameter comes from an analysis of the most recent National Health Interview Survey. All other weighting parameters are derived from an analysis of the most recent version of the U.S. Census Bureau’s American Community Survey.

The sample weighting is accomplished using an iterative proportional fitting (IPF) process that simultaneously balances the distributions of all variables. Weights are trimmed to prevent individual interviews from having too much influence on the final results. The use of these weights in statistical analysis ensures that the demographic characteristics of the sample closely approximate the demographic characteristics of the target populations.