Your Voice Carolinas is a survey center housed at the UNC Charlotte Urban Institute. Surveys can help us understand how individual perspectives and experiences for a large portion of the population and are often a key component of the Institute’s research projects. Survey data can also be combined with other types of data (administrative data, Census data, qualitative data, etc.) to help us understand quality of life across our 14-county region. In February 2026, Your Voice Carolinas will launch an annual survey to assess Charlotte regional residents’ perspectives on their personal wellbeing (Regional Wellbeing Survey).

A survey is a tool that researchers use to measure the attitudes and beliefs of a population of interest (in this case, the Institute’s 14-county region) by collecting responses from a sample of individuals that represent this larger group.

A “survey” and “poll” are similar in the sense that they both try to understand people’s opinions or experiences about a particular topic. The distinction comes from the intent behind the method.

Polls are typically focused on a specific public policy issue, with the intention of informing a wide audience about public opinion. Polls, such as those conducted during election campaigns, are administered repeatedly over shorter periods and focus on descriptive analyses.

With our survey, on the other hand, we are interested in understanding a more general concept of well-being and is framed around the various dimensions of quality of life as defined on our Charlotte-Mecklenburg and Carolinas Regional Explorer tools. The survey will be administered annually, and the Institute will conduct various descriptive and inferential analyses to generate a more nuanced understanding of well-being and quality of life across the region.

The concept of “quality of life” has not been measured extensively in the United States. When it has, it has been studied at the state or national, rather than the regional level. By partnering with the Happiness Research Institute, we hope to create a more well-rounded and effective measure of regional quality of life than anything that currently exists. Even better, our methods allow us to examine differences in quality of life across all 14 counties in the Institute’s region of service.

Survey data is one type of data we collect to tell the story of our communities and region. It is important to understand multiple types of data to understand an issue or opportunity

● Data on public perceptions and opinion (for example, the percentage of people that believe housing is our biggest problem)

● Data describing the extent and characteristics of an issue (for example, the number of housing units built in our region and the extent that supply is sufficient for current residents and those moving into the region)

● Data detailing people’s experience of an issue (for example, a story of a family’s experience from eviction to housing stability)

● Data on the historical and policy context of an issue (for example, the history of redlining that correlates with current housing and economic mobility outcomes).

There are numerous types of bias that can occur when conducting any survey, including sampling, non-response, and measurement biases.

Sampling bias occurs when the sample selected for the study is not representative of the larger population of interest. For example, a study of Mecklenburg County residents in which 80% of the sample reported household income of $150,000 or more per year would exhibit sampling bias, given that the county’s median household income is just under $85,000 per year. These survey findings would not be generalizable to Mecklenburg County residents as a whole.

Non-response bias occurs when individuals who respond to a survey differ fundamentally from those who do not. This type of bias typically occurs when a particular subgroup fails to respond to a survey and was a commonly discussed phenomenon after the 2016 presidential contest, when pollsters found (some) evidence that Republican-leaning voters were more likely to stop participating in polls than other members of the electorate. The resulting polls suffered from non-response bias, as they overestimated the likelihood that the Democratic candidate would win.

Measurement bias occurs when the wording of a survey skews results. This can stem from leading questions, loaded or emotional language, ambiguous or unclear wording, complex or double-barreled questions, or question ordering. In 2003, the Pew Research Center asked respondents whether they “favor[ed] or oppose[ed] taking military action in Iraq to end Saddam Hussein’s rule”. Sixty-eight percent favored military action, while 25% opposed it. When Pew added “even if it meant that U.S. forces might suffer thousands of casualties” to the end of this question, support dropped to only 43%, while 48% claimed to oppose it.

While it is impossible to eliminate all forms of bias resulting from a survey, the Institute is committed to ensuring our regional survey represents the viewpoints and experiences of our entire 14-county region by adhering to the following standards:

1. Utilizing a large sample size (N=4000) in order to increase the chances of residents being included (thus making it more likely that the sample will reflect the population as a whole).
2. Employing weighting techniques to correct for any bias that could still occur despite the large sample size.
3. Using both online and phone sampling to reach groups that are more likely to respond to either modality.
4. Engaging with well-established survey practices, ensuring that questions and the survey design are both valid and reliable.

The Charlotte Urban Institute is committed to ensuring that a broad representation of voices are heard in the Regional Wellbeing Survey. Thus, we are investing significant resources in the process to ensure that demographic groups that are typically less likely to respond to surveys participate.

The survey population includes all residents of the Institute’s 14-county region who are at least 18 years of age. In partnership with our selected vendor, the Institute will have access to two commercial panels (lists), one which consists of individuals within the region who have provided their email addresses to take online surveys, and one which consists of a list of phone numbers, also of individuals within the region, (cellphone and landline) that will be used to randomly select participants.

Yes, survey responses will be confidential. There will be no identifying information collected or stored from survey participants.