Sean Case

Data science and investigations using R

Recent posts

Sep 5, 2019
Ten years of Wisconsin votes - effect of voting machines My previous posts have considered correlations between voting machine use and the vote for Trump in 2016 in Wisconsin at the municipality and reporting unit level. In this post I will compare the results from 2016 to those in elections from 2008 to 2018, to see if the correlation of voting machines with the election results was a one-off, or if it is a common trend throughout the years. I also look at the results at county level.
May 1, 2018
Hunting for suspicious Twitter accounts with rtweet part 3 - Investigating user staggerlee420 The third part in a series of posts looking at how rtweet can be used to analyse Twitter data to try to identify suspicious accounts, i.e. bots or fake / sockpuppet accounts. Here we will look at one user identified as suspicious in part 2 - staggerlee420. This user was recently banned by Twitter. Using their user data and a sample of tweets collected in February 2018, we will try to determine whether this influential account was an automated bot, a fake / troll account, or just a regular user with strong opinions!
Mar 4, 2018
Hunting for suspicious Twitter accounts with rtweet part 2 - Identifying suspicious Twitter bot / fake accounts for further investigation The second part in a series of posts looking at how rtweet can be used to analyse Twitter data to try to identify suspicious accounts, i.e. bots or fake / sockpuppet accounts. Here we will analyse posting and account properties from our dataset of the hashtag 'firerosenstein' to find the most influential suspicious accounts.
Feb 18, 2018
Hunting for suspicious Twitter accounts with rtweet part 1 - Getting the Twitter data and preparing it for Gephi The first part in a series of posts looking at how rtweet can be used to analyse Twitter data to try to identify suspicious accounts, i.e. bots or fake / sockpuppet accounts. Here we will look at how to use the rtweet package to search for a hashtag, get out the tweet and user data, and then prepare the data for use in a network analysis programme such as Gephi, or within Twitter's own igraph package.
Oct 13, 2017
The effect of optical scanners on reducing turnout in Wisconsin The Optech Eagle optical scanner was recently decertified by Wisconsin Election authorities. Voter turnout in Racine county where this optical scanner was used (a county that was not part of the recount) was unusually low. Could this have been difference be due to intential manipulation, or was it simply a malfunction in the optical scanners? Do other counties see the same effect?
Sep 23, 2017
Republican vote 2016 with voting machines at reporting unit level The use of voting machines in Wisconsin is statistically significant in increasing the vote % for Trump at the reporting unit level. The size of the effect suggests that voting machine use could explain up to a 13,000 vote swing to Trump (he won Wisconsin by approximately 23,000 votes).
Apr 23, 2017
Influence of Sequoia AVC Edge touchscreen use on the Wisconsin Trump vote The proportion of votes coming from use of the Sequoia AVC Edge electronic voting system in Wisconsin was significantly correlated with increases in vote % for Trump in the 2016 Republican primary and the 2016 Presidential vote. This was not the case for the 2012 Presidential contest.
All articles →