Researchers often use averages to describe data. The average (or the mean) of a data set can be used to identify the central value of the group, or what is typical. While valuable, it’s also important to understand the range of data—the highs and lows. What might we miss by focusing on the average?
When considering averages, important questions to ask are: Is the data distributed normally creating a bell-shaped curve? Is the data skewed to one side leaving a tail at either end? When the data is skewed, the average is pulled to one side and is no longer located in the center; thus, the average would not be an appropriate representation of the typical value in the group. Even with normal distribution, considering the range of data values is imperative. For example, the state average drug overdose mortality rate may not look like a state problem compared to other state averages, but it could be a serious problem for some counties in the state.
In this brief, Range Matters: Rural Averages Can Conceal Important Information, the NC Rural Health Research Program uses three examples to demonstrate how focusing on averages without also considering the data range can conceal important information. The examples look at some conclusions drawn from commonly used indicators: 1) average rural hospital profitability, 2) distance from closed rural hospitals to the next closest hospital, and 3) HIV prevalence by county.