The shooting of Trayvon Martin was one of the biggest stories of 2012. But it didn’t start out that way — it began as just another local crime story, and without a number of key points of amplification along the way, it could have easily remained one. How did the death of a black Florida teen become what the Pew Research Center says was the most covered news story with a racial component in the past five years?

Three researchers at the MIT Center for Civic Media published a paper this week that tries to answer that qustion. “The Battle for ‘Trayvon Martin’: Mapping a Media Controversy On- and Offline” tracks how the event was covered immediately after Martin’s death, and over the following days and weeks as the story ascended from local broadcast news to national newspapers and the web.

The paper’s authors — Erhardt Graeff, Matt Stempeck, and Ethan Zuckerman — were specifically interested in determining how alternative and participatory media might have influenced the narrative arc of the Trayvon Martin story. The idea for the paper was sparked by a blog post Stempeck wrote in 2012. “This seemed like a good opportunity to look at a different kind of event,” Graeff said: “a crime story that had ballooned into a much larger set of issues.” As the authors put it:

Our primary research objective was to understand the relative prominence and importance of online and offline media at different points in the Trayvon Martin story. We also find a couple of general principles which may apply in stories beyond this specific case. Namely, that broadcast media is still important as an amplifier and gatekeeper, but that it is susceptible to media activists working through participatory media to co-create the news and influence the framing of major controversies.

To achieve this goal, the authors relied on a number of distinct datasets. To measure engagement on social platforms, they looked at individual tweets and the hashtags that united them. To monitor TV coverage, they pulled in closed-captioning transcripts. They tracked Google searches for both “Trayvon Martin” and “George Zimmerman,” his killer. In addition to measuring frequency of appearances in print media, they also measured front-page appearances “as a percentage of total physical column-inches on the front page of The New York Times, Los Angeles Times, Washington Post, and New York Post.” Looking at the click rates on Bitly links helped them understand the role that race-specific media played in the spread of the story. They also compared those datasets with the growing number of signatures on a Change.org petition that begin circulating on March 8, 2012.

But the central tool in executing the project was Media Cloud, a project from Harvard’s Berkman Center for Internet and Society that collects and analyzes both mainstream and alternative, digital news streams — its functions include “media definition, crawling, text extraction, word vectoring, and analysis.” Out of the over 27,000 sources Media Cloud draws on, 1,570 were relevant to the timeframe of the research question, February 26 to April 30, 2012. (The research didn’t deal with Zimmerman’s trial or its coverage in the media.) From those, Media Cloud ultimately found 359 sources, with a total of 5,665 stories, that were relevant to the story following Martin’s death.

Prior to this research, Media Cloud was used by Harvard’s Yochai Benkler, one of its designers, to track media coverage of the SOPA-PIPA debate and to map that controversy via links throughout the media ecosystem. Benkler found that digital media like Reddit, Techdirt, and AmericanCensorship.org “were the most influential sources in the media ecosystem as ranked by incoming links, overshadowing the impact of traditional media sources.” But as the authors point out, Benkler’s subject matter was inherently Internet-centric; with this paper, they sought to repeat the controversy mapping on a less natively digital subject.

The researchers present their findings by breaking the Trayvon Martin story into five acts. The first spans February 26 to March 6, during which time the story did not make it beyond of Florida media. Local TV news covered it, followed in the following days by stories in the Orlando Sentinel and The Miami Herald. “The news story, initially framed as a fight between two people in an area known for occasional violence, stood little chance of attracting significant media attention,” the authors write.

The second act is where we first meet Benjamin Crump, the civil rights attorney Martin’s family retained, and the publicist Crump hired, Ryan Julison. It was through Julison’s efforts that national news outlets, including Reuters and CBS, got wind of mounting concerns over details in the case — for example, why was Zimmerman carrying a gun? From there, the paper argues, digital platforms like The Huffington Post and the Black Youth Project began to amplify the story, but with the addition of significant misinformation.

It’s also in this stage that the circulation of the Change.org petition began to grow exponentially. With 217 signatures its first day, it grew to 2,492 the next day, picking up thousands a day before nearly reaching 13,000 on March 13. The paper also notes that celebrities sharing the petition on Twitter caused a 900 percent spike in traffic between March 12 and 15.

The third act, from March 16 to 22, saw the release of the audio file documenting the last few moments before Martin died. That audio provided what the authors call an “actuality” for TV and radio broadcasters to build a story around, which is why mainstream media is the central focus of this phase.

At the same time, protests started popping up in Florida, New York City, and London, providing “actualities” for newspapers, leading to the first national front-page story about Martin on March 22.

Then came the peak in Martin coverage in Act IV, from March 23 to April 10. It was during this period that President Obama weighed in with his comment on racial profiling. After that, Graeff says, “We see this kind of open season for political actors trying to use the attention dedicated to this story for political gain.” On the left, activists tried to use Trayvon Martin to advance their gun law legislation.

Then, conservative bloggers at sites like Wagist.com sought to push back by searching for evidence of “a troubled youth” in Martin’s social media profiles. This pursuit eventually led to The Miami Herald publishing a story describing an incident in which Martin was suspended for carrying a bag of marijuana. Though conservative bloggers could not prove Martin was a drug dealer, they were successful in getting the mainstream media to address that claim. This led liberal, digital media such as ThinkProgress and Gothamist to complain about the tactic “at the level of metanarrative, covering the coverage of the claim.”

Finally, in Act V, the authors argue that pressure from across the media ecosystems led authorities in Florida to take Zimmerman into custody. This once again provided the news hook necessary to spark a mainstream news cycle, with front-page coverage peaking the day after his arrest. Simultaneously, Google searches for “George Zimmerman” reached a high, as those who had managed to ignore the story previously sought context.

“Cable news channels FOX News and MSNBC aired a series of character attacks, while HLN’s Nancy Grace weighed in on whether or not Zimmerman cried in his jail cell,” the authors write. The spike in broadcast coverage, it’s important to note, outlasted Google searches, Media Cloud mentions, and newspaper front pages. “Like Hollywood’s penchant for sequels to popular film franchises, it’s possible that once a story and its characters have been introduced, it’s relatively frictionless for TV news programs to return with greater frequency to the story,” write the authors.

What can observers of media learn from this story? “We realized that the broadcast media were key to driving attention to this story, which is counter to the popular narrative that we’ve been putting together over the last few years about the networked public sphere,” Graeff says.

But at the same time, they learned broadcast and the rest of the mainstream media are highly susceptible to the agenda-setting of digital and alternative media: “We believe the national attention brought to the story through broadcast media allowed groups like the Black Youth Project to amplify stories to their online communities, and informed actors like [Kevin] Cunningham who launched campaigns like the Change.org petition,” the researchers argue.

They also believe that digital activists are getting better at manipulating the media, promoting their own agendas and challenging official narratives. Newspapers, in turn, are vulnerable to this type of direction because of their reliance on newsy “actualities” like protests or other events.

They also found the extent to which communication on social platforms became a part of the broader media narrative significant — for example, when Howard University students released their “Am I suspicious?” video on YouTube. Write the authors:

In our study, social media’s role in agenda-setting is more complicated. Attention to the Martin killing comes initially from professional news outlets. However, online communities demonstrate agenda-setting power both by organizing protests like the Million Hoodie Marches and by influencing online dialog, suggesting alternative interpretations of events.

Our work suggests a mechanism through which social media users introduce potentially deviant frames into the mainstream: they harness ideas to a high attention story already underway and attempt to direct the attention generated by the story towards their interpretations and views.

Going forward, this study provides an interesting foundation for thinking about how our media are interrelated, and how various facts, anecdotes, and bits of misinformation make their way to the public.

“Can we start exploring the data not from identifying these topics of keywords upfront, but asking an algorithm to surface some of those for us?” asks Graeff. “What are some unusual things or clusters of news stories that will allow us to get a sense of news stories that otherwise wouldn’t be seen?”

In the future, Graeff says he’d like to be able to better track geography and know where stories are being published and talked about. The team is also interested in using natural language processing to track the spread of quotations from source to source. In addition, “automated coding and sentiment analysis” could be used to better understand how perspectives in the newsroom are molding stories — tools like OpenGender Tracker.

Ultimately, the goal is to create a suite of tools that activists, journalists, and academics can learn from. Says Graeff: “A lot of what we show here is that there are better methods for studying the media as a so-called media ecosystem that allow us to really understand how a story goes from barely a blip to a major national/international news event, and how controversies circle around that.”