Question:
If the 2018 election for U.S. senator from Florida were being held today, how would you vote if the candidates were…

Bill Nelson, the Democrat 48%
Rick Scott, the Republican 42%
Someone else 2%
Don’t Know 7%

About this Poll
The University of North Florida (UNF), Florida Statewide Poll was conducted by the Public Opinion Research Lab (PORL) at UNF Monday, January 29, through Sunday, February 4, by live callers via the telephone, and calls were made from 5 p.m. to 9 p.m. Monday through Friday, 12 p.m. to 9 p.m. Saturday and 11 a.m. to 2:30 p.m. Sunday. The sample of phone numbers was created through the voter file provided by Florida’s Division of Elections September 2017 update and selected through the use of probability sampling among Florida registered voters in the Florida voter file. Interviews were conducted in English and Spanish by UNF undergraduate and graduate students. Overall, there were 619 completed surveys of Florida registered voters, 18 years of age or older.

To determine likely voters, respondents were asked two questions related to the upcoming midterm elections. First, they were asked on a scale of one (never) to five (always), how often do they vote in non-presidential elections. Second, respondents were asked about their enthusiasm for the upcoming election (not at all, somewhat or very enthusiastic). Respondents who answered four or five for the frequency of voting question, or answered a three to the frequency of voting question and also answered somewhat or very enthused about the upcoming midterm elections were included as likely voters.

The margin of sampling error for the total sample is +/- 3.9 percentage points.The margin of error for likely voters in the November 2018 midterm election is +/- 4.7 percentage points. This study had a 13.8% response rate. The American Association of Public Opinion Research (AAPOR) Response Rate 3 (RR3) calculation was used which consists of an estimate of what proportion of cases of unknown eligibility are actually eligible. The breakdown of completed responses on a landline phone to a cell phone was 27 to 73 percent. A single interviewer, through hand dialing, upon reaching the specific registered voter as identified in the Florida voter file, asked the respondent to participate, regardless of landline telephone or cell phone.

To ensure a representative sample being collected, the state was stratified using the 10 Florida designated market areas (DMA). DMAs are defined as regions where the population can receive the same or similar television and radio station offerings, as well as other types of media including newspapers and Internet content. In addition, because of Miami-Dade County’s unique population, it was separately accounted for in its own strata, creating 11 strata from the 10 DMAs. Quotas were placed on each of these stratified areas to ensure a proportionate amount of completed surveys from across the state. Data were then weighted by partisan registration, gender, race, age, and education. Education weights were created from the 2016 American Community Survey (ACS). Partisan registration, gender, race, and age weights were created from the September 2017 update of the Florida voter file to match the active registered voters in the state of Florida. These demographic characteristics were pulled from the voter file list. All weighted demographic variables were applied using the SPSS version 23 rake weighting function. There were no statistical adjustments made due to design effects. This survey was sponsored by the UNF PORL and directed by Dr. Michael Binder, UNF associate professor of political science.