Excel and Google Sheets can also be used to explore correlation coefficients and the distribution of error terms from bivariate regression analysis. The following example illustrates the relationship between unemployment and percent of the population without health insurance. It was discovered that a moderate positive correlation exists between unemployment and percent without health insurance by U.S. state. The model underestimates the percent uninsured for Texas, and over estimates percent uninsured for Massachusetts.
Another example, shown below, explores the association between smoking rates and average number of physically unhealthy days per month.