SurveyUSA is an independent, non-partisan, apolitical research company that conducts opinion surveys for media, academic institutions, commercial clients, non-profits, governments, agencies, and elected officials. SurveyUSA opinion research is conducted using a methodology optimized for each particular project. In some cases, this means data is collected 100% by telephone; in some cases, 100% online; and in other cases, a blend of the two. For those projects that are conducted “mixed-mode” (or “multi-mode”): Respondents who have a home (landline) telephone are interviewed by phone, sometimes using live interviewers, other times using the recorded voice of a professional announcer. The youngest male is requested on approximately 30% of calls to home phones, the youngest adult is requested on approximately 70% of calls. This method of intra-household selection reduces the potential for age and gender imbalance in the unweighted sample. Re-attempts are made to busy signals, no-answers, and answering machines. Landline telephone sample is purchased from Aristotle of Washington DC. Respondents who do not use a home telephone are interviewed on an electronic device, which means, for some projects, that call-center employees hand-dial cell phones and interview respondents verbally on the respondent’s cell phone, and means, for other projects, that SurveyUSA displays the questions visually on the respondent’s phone, tablet, or other device. Sample for respondents who do not use a home telephone is purchased from Aristotle or from one of several other research companies that provide access to cell respondents. Where meaningful, SurveyUSA indicates the percentage of respondents who use a home phone and the percentage who do not, and crosstabs by this distinction. Responses are minimally weighted to U.S. Census or voter file targets for gender, age and race. Target (cell) weighting is used. On questionnaires that ask about political party identification, SurveyUSA may or may not weight to Party ID, depending on client preference. If a region crosstab is shown, the geographies included in each region are defined here. Where necessary, questions and answer choices are rotated to prevent order bias, recency, and latency effects. On some studies, certain populations are over-sampled, so that the number of unweighted respondents exceeds the number of weighted respondents. Each individual SurveyUSA release contains the date(s) on which interviews are conducted and a release date. If interviewing for a particular study is conducted in Spanish, or in any other foreign language, it will be noted on the specific release. If no notation appears, interviews are conducted in English. Where respondents are filtered, such as adults, filtered to registered voters, in turn filtered to likely voters, SurveyUSA describes the filtering on the specific release. On pre-election polls in geographies with early voting, SurveyUSA differentiates between respondents who have already voted and those who are likely to vote but have not yet done so. On research completed prior to 12/31/16, SurveyUSA assigned to each question within the instrument a theoretical margin of sampling error. Effective 01/01/17, SurveyUSA assigns to each question within the instrument a credibility interval, which better reflects the sampling uncertainties associated with gathering some percentage of respondent answers using non-probability sample. Though commonly cited in the presentation of research results, “sampling error” is only one of many types of error that may influence the outcome of an opinion research study. More practical concerns include the way in which questions are worded and ordered, the inability to contact some, the refusal of others to be interviewed, and the difficulty of translating each questionnaire into all possible languages and dialects. Non-sampling errors cannot be quantified. This statement conforms to the principles of disclosure as recommended by the National Council on Public Polls (NCPP). Questions about SurveyUSA research can be addressed to email@example.com.