DataCamp: Analyzing Election and Polling Data in R Review

Analyzing Election and Polling Data in R is a new introductory level R course released on DataCamp just head of the November 2018 midterm elections that uses US and UK voting data to teach concepts in R programming. The course focuses on data manipulation with dplyr, plotting with ggplot2 and linear regression with base R and also touches on working with dates, time series data, plotting on maps and confidence intervals. It is recommended to have taken an introductory R course and have some exposure to the Tidyverse package in R before taking this course. Analyzing Election and Polling Data in R is a paid DataCamp course: the first chapter is free but the other three require a subscription to access.

Analyzing Election and Polling Data is divided into 4 chapters. The first section explores President Trump's approval rating over time while introducing some basic dplyr and ggplot functionality and rolling averages. Chapter 2 explores generic ballot polls and covers visualizing time series, calculating error in polls, plotting error bars and fitting a linear model. The third section explores 2016 election data and Brexit data and touches on merging data, creating county-level maps and multivariate regression in R. The final discusses making predictions from regression models in the context of the 2018 midterm elections and 2020 presidential election.

Each chapter has roughly 15 video lectures and programming exercises exploring election data with R. The lecture videos consist of slide show presentations of key concepts and code with voice overs by the instructor; unlike most other DataCamp courses the instructor does not appear on screen in this course. Since your attention should be focused on the information and code in the slides anyway, this is not a great loss, although some brief face time at the beginning and end of the course might be nice. The programming exercises are relatively short and straightforward, with most only requiring you to fill in a couple key arguments into code skeletons that are mostly written for you in advance. If you have some exposure to R and the tidyverse, you should be able to get through the course in less than the prescribed 4 hour completion time without taking too many hints.

Analyzing Election and Polling Data provides a gentle overview of a variety of topics in R using political data as motivating examples. As an introductory level course, the analysis doesn't get particularly complicated and focuses more on using the data as a backdrop for teaching R than diving deep into any particular data set. A project that takes the concepts from this course and goes deeper with a specific data set would be a good complement to this course.

I give Analyzing Election and Polling Data 3.75 out of 5 stars: Good.