What do Trump’s tweets tell us? Is there a pattern or just poor grammar?
A lot has been said about the way Trump tweets. From ‘covfefe’ to capitalizations that can give you a headache — everything that comes through the handle @realdonaldtrump has been broken down, dissected and decoded by the media. We are all just trying to make sense of who is Tim Apple and if he thinks DACA (Deferred Action for Childhood Arrivals) is a real person.
As journalists, the top advice we are given is report based on data and not anecdotal evidence. Keeping that as my guiding light, I extracted Trump’s tweets dating back to May 2018 to look for a trend and achieve some understanding of his actions. The Twitter data was nothing short of a treasure trove.
I first picked my favorite pet peeve out of the list — his overuse of the words “millions and billions.” I have come across many tweets by Trump that have had the word million or billion in it. But again, does the data support my belief? Considering almost two years' worth of tweets, words million and billion have been used over 200 times.
Exactly 15 tweets mention the word ‘impeachment’ and almost 200 tweets talk about ‘China.’
But is there a trend? Does the President plug in these words more often during a certain period than the other?
The data does tell a story: In the month of July 2018, Trump gave out “billions of dollars” in emergency relief to farmers hurt by trade tariffs, which explains the 21 tweets with the term ‘millions’ or ‘billions’ in them.
China and impeachment — which also happen to be the prevailing news topics— are mentioned incessantly during the closing months of 2019.
This analysis was performed using Natural Language Processing in Python. The data files along with the open-source code can be found in this GitHub link, if you would like to perform your own analysis.
Intrigued to know more? Here’s another revelation about Trump found in the Reddit dungeons. A graph by Dan Snow, which shows how Trump’s tweets per hour increase each time Fox & Friends airs. Side note: This is just a fun fact not a direct correlation.
Note from the Editor: This post is part of a series, written by students of the Spring 2019 Data Journalism course in the Newmark Graduate School of Journalism at CUNY, in which they share their work and thought process. Each week we have a Data Fest in which two of the class reporters present a data set, along with a brief critique and overview of what they did and discovered.
