About the Data

This website, and the data featured on it, are a product of State of Health — an initiative of the Center on Society and Health at Virginia Commonwealth University, funded by the Robert Wood Johnson Foundation. This site focuses on mapping the results of health outcomes and key conditions that shape health. As part of this project, our team also conducted extensive statistical testing to determine which variables were the best predictors of state health outcomes. The method we used, called best subset multiple regression, is described in the technical report to the right.

That analysis identified eight variables that performed best in predicting health outcomes at the state level. Those variables were (1) median household income, (2) the proportion of the population with a bachelor’s degree or higher, (3) neighborhood resources for children, (4) childhood trauma, (5) avoidable hospitalizations, (6) access to primary care physicians, (7) state income support, and (8) tobacco tax rates.

These associations do not necessarily imply causal relationships — in other words, what predicts health does not always directly affect health — nor do these correlations provide evidence about the effectiveness of specific policy solutions. What’s important to note about these variables are their interrelationships. For instance, states with fewer neighborhood resources for children also tend to have lower median household income, fewer college graduates, less access to primary care, more food insecurity, more avoidable hospitalizations, greater childhood trauma, and more adults in prison. People in these same states also had more activity limitations and higher rates of stroke, lung cancer, breast cancer, adult overweight/obesity, infant mortality, motor vehicle fatalities, and smoking, as well as higher overall mortality rates and lower life expectancy. Download our data to explore similar interrelationships with the other seven variables.

We hope you’ve found this summary useful. Please reach out (societyhealth@vcu.edu) with any questions, suggestions for how to make the site work even better for you, or to share wonderful stories about how our data has helped you move health-supportive policies and programs forward in your state or across the country. We’re excited to hear from you!