Methodology

AAPOR TI_CharterMember

 

The Rutgers-Eagleton Poll conducts statewide telephone surveys of the New Jersey adult and New Jersey registered voter populations with a scientifically selected random sample using a random-digit-dialing (RDD) method. Respondents within a household are selected by asking randomly for the youngest adult male or female current available (18 or older). If the named gender is not available, the youngest adult of the other gender is interviewed. The poll is available in Spanish for respondents who requested to do it in that language.

In each New Jersey statewide poll, our goal is to reach at least 60% of our respondents on cell phones, to best reflect the estimated cell phone-only household rate in New Jersey based on CDC estimates. This number will increase depending on details of a specific poll and as over time the number of cell phone only households increases. All Rutgers-Eagleton polls are conducted using live interviewers.

Our samples are typically balanced to match parameters for sex, age, education, race/ethnicity, region and phone use. The sex, age, education, race/ethnicity and region parameters are derived from American Community Survey PUMS data. The phone use parameter is derived from estimates provided by the National Health Interview Survey Early Release Program.

Weighting is done in two 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. This adjustment also accounts for the overlapping landline and cell sample frames and the relative sizes of each frame and each sample. The second stage of weighting balances sample demographics to match target population parameters. This weighting is accomplished using SPSSINC RAKE, an SPSS extension module that simultaneously balances the distributions of all variables using the GENLOG procedure. Weights are trimmed to prevent individual interviews from having too much influence on the final results.

All surveys are subject to sampling error, which is the expected probable difference between interviewing everyone in a population versus a scientific sampling drawn from that population. Sampling error should be adjusted to recognize the effect of weighting the data to better match the population. Thus if our simple sampling error is +/- 3.6 percentage points and 50 percent of New Jersey registered voters favored a particular position, one would be 95 percent sure that the true figure is between 46.4 and 53.6 percent (50 +/-3.6) if all New Jersey registered voters were interviewed, rather than just a sample. We also account for the design effects of our samples and report an adjusted margin of error that reflects these design effects.

Sampling error increases as the sample size decreases, so statements based on various population subgroups are subject to more error than are statements based on the total sample. Sampling error does not take into account other sources of variation inherent in public opinion studies, such as non-response, question wording or context effects.

The Rutgers-Eagleton Poll is generally fielded by the Eagleton Center for Public Interest Polling. The questionnaire is developed and all data analyses are completed in house. The Rutgers-Eagleton Poll is paid for and sponsored by the Eagleton Institute of Politics, Rutgers University, a non-partisan academic center for the study of politics and the political process. Full questionnaires are available on request, and can also be accessed at the Eagleton Poll archive at www.eagletonpoll.rutgers.edu. Please see individual press releases for more specific details, or for more information, please contact akoning@rutgers.edu.

To learn more about the AAPOR Transparency Initiative, see here.