How well does a value-based regression model perform in the 2016 Presidential Election?

Using multiple regression to investigate values predictive of Donald Trump voters in the 2016 election

Zach Alexander
Towards Data Science
6 min readDec 30, 2019

--

Comparison of regression model performance against actual U.S. presidential election results in 2016.
A D3.js map comparing the actual 2016 election results to the model predictions. Visit www.zach-alexander.com/pres2016-regression to interact with the full visualization.

For some, it may be hard to believe, but the 2020 presidential election is roughly 300 days away. In what is shaping up to be another contentious…

--

--

Data Engineer at the Center for Court Innovation. MS in Data Science — CUNY