The Boston bombings gave us a glimpse into the contentious world of Twitter news-breaking. Vigilante investigators took to Twitter and Reddit to find the bombers, and soon the names “Sunil Tripathi” and “Mike Mulugeta” began clanging around the information sphere.
As Alexis Madrigal highlighted in the Atlantic, that intel was totally wrong, of course, but the episode did demonstrate a phenomenon that now happens some 20% of the time, according to a new study from the University of Edinburgh’s School of Informatics: While traditional newswires still dominate the dissemination game, Twitter still scoops old media on certain stories—largely ones dealing with sports, major disasters, and sometimes riots.
This table measures the Twitter vs. Newswire lead time on news stories between late June of 2011 and mid-September of that year (the faster newsbreaker is in bold):
While newswires had the majority of scoops, Twitter still broke the news on the England riot mortalities by as much as an hour ahead of the wire. Meanwhile, 95% of newswire stories also made their way onto the microblogging platform.
Miles Osborne, lead author of the study, says that the exchange between Twitter and newswire lead times could mean that the reporting cycle’s due diligence shrinks over time—and anecdotally, we see that it already has in some cases. Still, Osborne’s data showed that the newswire is still the main force in information sharing, and he doesn’t anticipate that changing.
“My prediction is that … news services will simply use Twitter as another dissemination vector,” Osborne told Co.Exist in an email. “In my opinion the broad findings will continue. For example, news of some border treaty will never surface first on Twitter. Likewise, someone being robbed in downtown Boston might well appear first on Twitter.”
Or, Twitter could lead in breaking news of events that might otherwise be swept under the rug. Osborne’s study, which is being presented at the International AAAI Conference on Weblogs and Social Media (ICWSM) in Boston this week, didn’t measure how many acts of political dissent were broken on Twitter first, but his list of sample stories with a Twitter lead time includes Malaysian police using tear gas on protesters:
The findings were no small undertaking. While Osborne’s study relied partially on algorithms to determine tweet clusters surrounding the news, it also required manual sorting through a massive screen of tweet diarrhea. Out of the 1,000 unique Twitter events Osborne’s team identified, only 54 were determined to be signals in the noise. “Noise,” Osborne explained, were the stories that weren’t actually stories—Justin Bieber quotes, he said, were an example.