The survey results are based on interviews conducted via the PPD National Internet Polling Panel. Interviews are NOT conducted on most national holidays and samples can include various fractions/percentages of repeat interviews from panelists, which we have demonstrated provides a more accurate gauge of shifting public opinion.
All responses in the sample should be treated as “opt-in Internet panel” even though a percentage of respondents were specifically targeted based on registration status (more on that below in population). They are still ultimately considered opt-in and we do NOT treat them as a random sample. Respondents either received a mailer, an email, responded to social media ads or a website-based ad, or contacted us through another mode.
During screening or initial interviews, respondents are asked to give their names, contact information (i.e. email and/or phone), as well as the city and zip code where they are registered or plan to register to vote. This is so we can attempt to re-interview, to obtain regional data etc. The information is also used to verify registration status when possible (based on the state) for respondents who were not already identified.
We’ll attempt to contact a respondent for a repeat interview up to 10 times before removing them from the panel.
The top line results are of likely voters, or at least our best estimate of registered voters we view to be most likely to vote based on past voting history, enthusiasm and registration status. We do not includes respondents who report they are not registered during initial interviews in the results, but also don’t immediately remove them as a potential panelist. They may register in the future and we view them as worthy of a follow-up.
In the past, we’ve found follow-up greatly reduces the probability of unintentionally excluding new voters of various ages.
If you’d like to inquire about participation, you can do so by contacting us through email here. We are looking to expand our panel with those who are willing to conduct repeat interviews, are of age, eligible to vote and are qualified via a screening questionnaire. Include contact information, as well as zip code and/or city and state.
The PPD Poll does not weight based on party affiliation (party ID). Polling results are weighted based on the U.S. Census Current Population Survey for demographics such age, sex, race, education and region. It uses a likely voter model based on responses to screening questions relating to prior voting history, enthusiasm and registration status etc.
The poll does oversample, but that’s primarily the result of the entire sample being of registered voters and the use of a likely voter model. Top line numbers do not report the results of those who answer “No, I do not” or “Not sure” to the question: “Do you expect to vote in the [upcoming election] this coming November, or not?” We will cycle back to these respondents and attempt to interview them again at some point during the following week. Often, they change their minds and, if they do, those results are included in the sample.
Because of how we screen, the disparity is not significant.
We cannot accurately estimate a margin of error (MoE) because the sample is not a probability sample that used a random selection. Though we don’t know exactly who will and will not respond to the panel for a given day, they are still treated as opt-in. Instead, we use a bootstrap method with a standard 95% confidence interval. Admittedly, this can be a bit difficult with more than one or two choices, but we have had success in accounting for this in the past. The 95% confidence interval varies daily and will be posted along with results.
To reduce human error, we have begun to use StatKey as a basis to calculate the 95% confidence interval for candidates’ shares of the vote. We do so by generating 5,000 responses with the weighted results for the average three-day results.
Whether we call a respondent to request or conduct a repeat interview, they take it online or via email, the questions are worded exactly as they are in the link below and vote preference choices are randomized. (Caller A = Trump is the first choice; Caller B = Clinton is the first choice; Caller C = Johnson is the first choice; Caller D = Stein is the first choice.)