Five Things to Know About County Profiles
Benefits and limitations of data profiles, dashboards & other data compilations

Tips to help you understand the strengths and weaknesses of data profiles, dashboards and other data compilations.

Author: Mike Cline, North Carolina State Demographer

Accessing data has never been easier. It’s an exciting time for data fans, like myself, to see the increasing availability and intuitiveness of data.  However, it’s also important to understand the strengths and weaknesses of data profiles, dashboards, and other handy data compilations. Keep these tips in mind when using these useful tools.

1) Profiles are an easy way to share key statistics of your area with decision-makers.

Profiles are an easy way to compare your area with others or with itself. For example, quickly discerning how your county ranks compared to other counties, neighboring counties, similar counties, and/or the state. Or seeing how an indicator has changed over time. Information is often presented in compelling and quick-to-grasp graphics, saving the user time.

2) Profiles are like peering into a community through a small window.

They are limited by the purposes for which the profiles were created; the resources of the organization providing the profiles; and/or data availability.

General purpose profiles are created to provide a resource for commonly asked questions. They serve as a great first stop for accessing data. Some popular examples in North Carolina are:  

Some profiles/dashboards may be created to focus on a key demographic group or activity, or to trace pre-determined key indicators. Example:

3) Any given indicator in a profile provides an answer but not a comprehensive answer.

For instance: Showing median household income can provide a general indicator about household incomes in your county – but it does not provide any information about income disparities. Does your county only have very high-income and very low-income households? How many low-income households? TIP: Good profiles provide source information so that you can dig into the data beginning with that source.

4) Data profiles of a selected geographic area can mask differences within that geographic area.

Profiles usually provide data for a specific geographic area such as counties and municipalities. Counties are the most used because their boundaries are generally stable over time, enabling you to compare indicators to historical periods.  Counties are also the smallest geography for which most administrative data are reported (e.g. vital statistics, vehicle registrations, etc.).1

Survey-based estimates, such as the American Community Survey (ACS), are also more reliable at the county level than estimates made for subcounty areas, such as census tracts (i.e. the margins of error are larger for subcounty areas).

However, county characteristics can mask differences within the county. Overall, 10% of the population of a county could live in households with incomes below poverty. Within the county, there could be one or two neighborhoods (or census tracts) where 90% of the population live in households with incomes below poverty. Geographers call this the modified area unit problem (MAUP). 

5) Profiles are only as good as their source data. 

“…the more the data, the surer we fool ourselves.” Xiao-Li Meng (reflecting on Big Data usefulness)

Pretty visualizations don’t make up for bad data. And data can be bad. Severe deficiencies can limit the usefulness of a dataset. Thankfully, most folks who prepare data profiles have spent time evaluating the datasets incorporated within their given profile. Still, no data are perfect and there are limitations to every dataset of which you should be aware. 

Good profiles will provide some background material that provides information about the limitations of any given dataset within the profile. Useful data profiles should provide a link to the original data source so that you can read about the methodologies, assumptions, and limitations of the data. The original source may also provide you with more detailed data on a specific subject. 

Along with the source data, it is important to know the date of the estimate and/or production date/vintage. For instance, NC OSBM produces annual population estimates AND projections. We revise our previous estimates and projections based upon new input data, corrections to previous input data, model improvements, and/or (for projections) updates to trends shown in new estimates (for instance more counties showing growth in the past year than experienced during the depth of the pandemic). 

Armed with this perspective, I encourage you to get out there and enjoy the data! 


1. More and more data are available for smaller geographies – sometimes down to the parcel. However, data are not always published for detailed geographies and are thus only available by request from the data owner.