4. It has been argued in a number of venues that a higher education level indicates a greater average income. The question for a college student might therefore be “Should I stay in school?”

a. Using publicly available U.S. Census data for the 50 states and Washington, DC, develop a linear regression model (causal forecasting) to see whether this relationship is true. (Note that some data massaging may be necessary.) How high was the R2 value (a measure of quality of fit)? Create a scatter plot of the data.

b. Does the relationship appear to be linear? If not, check a statistics book and try a nonlinear function.

How well does the nonlinear function perform?

c. Which five states have the highest average incomes, and which five states have the highest average education levels? From this study, do you believe that a higher average education level tends to cause a higher average income? Explain.

d. If you have studied (or will study) neural networks, using the same data, build a neural network prediction model and compare it to your statistical results.