Skip Header

Component ID: #ti138756772

Michael Binder, Associate Professor and Faculty Director-Public Opinion Research Lab at the University of North Florida

Data Story

The Public Opinion Research Laboratory (PORL) at the University of North Florida frequently uses American Community Survey (ACS) data tables in order to increase the accuracy of our state and local polls. By using ACS data to determine the age, sex, race, and educational breakdown of various counties in Florida, PORL is able to weight results so that the sample of individuals polled represents the total population being assessed.

Our most recent annual media survey conducted for the Florida Department of Transportation is an example of how these weights are created and what purpose they serve. Individuals in the state of Florida were asked a series of questions about their driving habits and opinions on highway safety, specifically concerning their awareness of the “Drive Sober or Get Pulled Over" safety message. This campaign was formed as a response to the rapid increase of fatalities caused by impaired driving nationwide. At the conclusion of the survey, we found that many more males had participated in the survey than females. In order to adjust for this, male responses were weighted down slightly based on the sex breakdown in the state of Florida so as to not over represent males, and thereby under-represent females.


Using ACS data to weight our survey leads to a greater confidence in the data that we collect, distribute, and report, as well as a time and cost savings for our sampling procedures. The ACS provides a baseline of what the population looks like at the national, state, and local level. Without these data, we would have no way of knowing if we had too many men or women in our survey, or too many adults with a college degree. I couldn’t imagine trying to gather data without some semblance of an idea of what the population actually looks like – and the Census and the ACS provide that. In many cases, the lack of proper weights can lead to inaccurate statistics, and therefore incorrect conclusions. 

In an era driven by data, it's of the utmost importance that the results we send out into the world are as accurate they can be, so we can provide precise information to inform decision-making. The Florida Department of Transportation (FDOT) uses the results from our media survey to better target their messaging in the upcoming year. For example, areas with the lowest awareness get increased messaging to help increase awareness of the “Drive Sober” campaign and dangers of drunk driving. Beyond geographic areas, FDOT uses this information to target specific demographics that may have lower awareness, such as males without a college degree.  Additionally, they use this information to reallocate resources. In the 2017 poll, we had very few people indicate that they had seen the “Drive Sober” message at sporting events and gas stations.  Because of that, advertising in those venues was reduced and directed toward areas that were more productive – such as electronic message boards over the highway and social media.



ACS Topics Used

Age, sex, race, and educational attainment

You May Be Interested In

Related Topics

Around the Bureau

Back to Header