A group project on Twitter and the 2016 US Election, submitted in partial fulfilment of the requirements for the degree of Masters of Science in Business Analytics.
This project acheived a second class honours grade one degree.
The 2016 US Presidential election was very different to previous campaigns. Prior to this election, traditional campaigning methods were used to gather and predict votes. In 2016, Nominee Hillary Clinton was forecasted to be triumphant with all polls pointing towards a comfortable victory over opponent Donald Trump. In the end, the polls proved to be a poor election predictor with Trump being triumphant. The polls got it so wrong, but how? Opinion polls can lead to sampling bias and lead people to give false statements and potentially lie. Social media eliminates these false predictors allowing people to express their true opinion publicly while also giving them the chance to remain anonymous. Twitter was the most used social media platform by both candidates, with Donald Trump utilizing it the most. It is the ultimate goldmine of all things opinionated. This project focuses on comparing both nominees Twitter accounts and public tweets during the final Election Day vs traditional public opinion polls. The results of this project showed that there is moderately positive correlation between the popularity of each candidate on Twitter and the number of times they were mentioned is user’s tweets on the final Election Day.