My AP Stat capstone sparked some interest, so I whipped up this model to predict average household income based on county socioeconomic data (taken from a Kaggle dataset using a Harvard census). Turns out that even if you use way more data than just unemployment rate—not to mention making it a census of the entire US instead of an in-state SRS—the r² is only 0.65. Interesting!
I have finals starting from tomorrow, but I spend my time doing this. Yay!