Skip to content

Topline Results from Oct. 2022 Times/Siena Polls in Arizona, Georgia, Nevada and Pennsylvania


Methodology

Sample

The survey is a stratified response-rate-adjusted probability-proportionate-to-size sample of the L2 voter file. The sample was selected by The New York Times in multiple steps to account for differential telephone coverage and nonresponse.

To adjust for noncoverage bias, the L2 voter file for each congressional district was first stratified by statehouse district, party, race, gender, marital status, household size, turnout history, age and homeownership. The proportion of registrants with a telephone number and the mean expected response rate were calculated for each stratum. The mean expected response rate was based on a model of unit nonresponse in prior Times/Siena surveys. The initial selection weight was equal to the reciprocal of a stratum’s mean telephone coverage and modeled response rate. For respondents with multiple telephone numbers on the L2 file, the number with the highest modeled response rate was selected.

Second, the probability of selection was weighted by the likelihood that a respondent would vote in the 2022 election, based on a model of turnout in 2010, 2018 and 2021 as a function of the respondent’s political and demographic characteristics.

Finally, the Nevada voter file was split between the First District and the rest of the state. The two Nevada samples were fielded and weighted separately.

Fielding

The sample was stratified by party, race and region and fielded by the Siena College Research Institute, with additional field work by ReconMR, IPOR at Roanoke College, and PORL at the University of North Florida. Interviewers asked for the person named on the voter file and ended the interview if the intended respondent was not available.

The instrument was translated into Spanish by ReconMR. Bilingual interviewers began the interview in English and were instructed to follow the lead of the respondent in determining whether to conduct the survey in English or Spanish. Monolingual Spanish-speaking respondents initially contacted by English speakers were recontacted by Spanish-speaking interviewers.

Weighting

The survey was weighted by The Times using the R survey package. Survey weights were trimmed at the 99th percentile.

The samples of Arizona, Georgia, Pennsylvania, Nevada’s First and the rest of Nevada were each weighted as follows:

First, the sample was adjusted for unequal probability of selection by strata.

Second, the sample was weighted to match voter-file-based parameters for the characteristics of the likely electorate. The sample was also weighted to match census-based estimates for the educational attainment of the likely electorate.

The estimates for the characteristics of the likely electorate were based on multiple models. First, a model estimated the likely turnout by state, based on recent turnout and the competitiveness of the midterm race. Second, an individual-level model of turnout in 2010, 2018 and 2021 was used to estimate the probability that registrants would participate in the midterm election as a function of their demographic and political characteristics. Finally, the individual-level estimates were adjusted to match the expected turnout by state.

Third, the weighted samples of Nevada’s First District and the rest of Nevada were combined and the samples balanced to ensure their appropriate state of the likely electorate.

Fourth, self-reported vote intention was incorporated into the estimate of a respondent’s likelihood to vote. That was based on a model of validated turnout in Times/Siena surveys since 2016 as a function of self-reported vote intention and the pre-survey modeled turnout probability, including an adjustment for the higher turnout of survey respondents than nonrespondents.

Fifth, the final weight for the likely electorate was equal to the initial likely electorate weight, multiplied by the final turnout probability incorporating self-reported vote intention divided by the initial modeled turnout probability.

Parameters

The following voter-file-based targets were used to weight the sample to match the characteristics of registered voters and the likely electorate:

• Party (NYT classifications based on L2 data and, in states without party registration or primary vote history, a model of partisanship based on previous Times/Siena polls)

• Age (self-reported age or voter file age if the respondent refuses)

• Gender (self-reported gender or voter file gender if the respondent refuses)

• Race (L2 model)

• Marital status (L2 model)

• Homeownership (L2 model)

• Region (NYT classification)

• Turnout history (NYT classifications based on L2 data)

• Vote method in the 2020 elections (NYT classifications based on L2 data)

The following census-based targets were used to weight the sample to match the characteristics of registered voters and the likely electorate:

• Educational attainment (NYT model based on A.C.S. and C.P.S. data)

• White/nonwhite x four-year college graduation (NYT model based on A.C.S. and C.P.S. data)