Massive volume of tweets directed at #MarchForOurLives activists
Two prominent #MarchForOurLives activists have been on the receiving end of a massive quantity of tweets since the February 14th mass shooting at Marjory Stoneman Douglas High School in Parkland, FL. Ed Summers, Bergis Jules from Documenting the Now and I analyzed tweets that mentioned the Twitter handles of Emma González and David Hogg. Emma and David are students at Marjory Stoneman Douglas High School and survivors of the mass shooting.
People from many different communities, both supporters and detractors, were involved in conversations surrounding the accounts of these two young activists. The communities included large verified accounts, media and journalists, politicians and various celebrities. Also present was a small group of high-volume accounts, some of which were created on or after the day of the shooting, February 14, 2018.
This collaboration includes analyses of two different datasets: a larger dataset containing 1.2 million tweets (10 days) obtained via Twitter’s Search API using twarc by Ed Summers and Bergis Jules and a smaller sample of 92,325 tweets (5 days) downloaded using Tweet Archivist which I used to create several network visualization graphs with Gephi.
1.2 million tweets were directed at both @Emma4Change and @davidhogg111 within a period of 10 days, from February 18 to February 27. Emma and David received 83.3 tweets per minute from 404,506 accounts.
- 80% of the 404,506 accounts only sent 3 tweets or less
- 13,991 (1.2%) of the tweets were sent by accounts created on or after February 14, 2018
- These 13,991 tweets were sent by 3,740 users (0.9%) of which 19 have been suspended, deleted or changed screen name
- 1,235 of those 13,991 tweets contained links out to the web. Ed Summers unshortened them and found 169 different URLs
- 167 users (0.04%) sent more than 100 tweets to either David Hogg or Emma Gonzalez over the 10 day timeframe
I started downloading tweets from Tweet Archivist on February 27 at 4pm EST and made user-to-user network graphs at three different intervals. Nodes represent Twitter users and edges are the connections between them: tweets, retweets and mentions.
3,000 tweets from February 27
The first visualization is a combined network graph of 3000 tweets from February 27; it shows 1500 tweets mentioning Emma’s account and 1500 tweets mentioning David’s account. Even though it’s a small sample, the difference in the activity between the two accounts is already noticeable. Emma had a large group of people retweeting her tweets — David did too, you can see the burst just to the left of his handle. But David also had an onslaught of trolls tweeting at him due to several large accounts mentioning his user name including Alex Jones, Thomas Wictor, Ryan Fournier, and an account called “Educating Liberals” (handle: Education4Libs), which is run by a YouTuber from Minnesota named Dylan Wheeler.
30,464 tweets from February 27 to February 28
I made a second combined graph of 13,497 tweets mentioning Emma4Change and 16,967 tweets mentioning davidhogg111; a total 30,464 tweets. I enlarged labels of some of the prominent accounts in the graph to show the complexity of these networks. Some of the labels are large, verified accounts of media, journalists, celebrities and politicians but there are also some trolls that were tweeting mostly at David. Other accounts such as CNN, NRA, DICKS and the POTUS account appear in this graph because they were included in tweets with one or both activists’ Twitter handles.
92,325 tweets from February 27 to March 2
I created two separate network graphs for each activist for the entire 5 day timeframe. The following graphs contain 46,277 tweets mentioning davidhogg111 and 46,048 tweets mentioning Emma4Change.
The differences between the graphs are interesting. I processed each graph exactly the same way using the same force-directed layout algorithms and both networks contain nearly the same amount of tweets, yet Emma has a solid mass surrounding her account and David’s account appears as if it was being pulled in all directions. There are clusters of supporters in each network graph however, David was on the receiving end of brigading from trolls, mostly due to conspiracy theories claiming David was a “crisis actor” which were widely spread by right wing websites and Twitter fights with Alex Jones.
While we were working on this blog, someone on Twitter asked if we could estimate the percentage of support versus trolling that was happening in these tweets, and we are really not sure how to measure that metric. We also discussed if it was possible to measure the impact of the small group of high-volume accounts that were tweeting hundreds of tweets per day at these two activists, but we are not sure how to quantify the effects of that activity either.
My personal opinion is they definitely have some impact, otherwise why would the operators of those accounts bother to do that. I’m not sure whether to label those accounts as bots, cyborgs or trolls but I think their activity can be categorized generally as SPAM.