Investigating England’s summer of disorder: The Guardian’s ‘Reading the Riots’ project
It all started on 4 August with a man shot dead by the police during an attempted arrest in north London. The death of 29-year-old Mark Duggan ignited a series of urban riots, looting and disorder, that the U.K. was no longer used to since the 1980s Brixton Riots. Two days after Duggan’s death, violence erupted from what had started as a peaceful protest demanding justice and an investigation into the dynamics of the shooting.
The 2011 U.K. riots crippled the country and attracted international media attention for four consecutive days. They were met by an unprecedented swift and harsh response on the side of British courts. The political establishment blamed the role played by social networks and instant messaging technology in spreading the violence across the country.
The Guardian’s 'Reading the Riots' is a multimedia project that applies data journalism in order to try to understand the reasons and individual stories behind such extraordinary events, to analyze the root causes and the consequences.
The project is the result of an ongoing cooperation between the Guardian and the London School of Economics (LSE). It is strongly focused on researching the social base of the riots. Divided into two main phases, 'Reading the Riots' combines print and online articles, video interviews and features, interactive maps and visualizations, and a series of open debates organized in the local communities affected by the riots.
"The first phase of Reading the Riots was completed in three months using confidential interviews with hundreds of people directly involved in the riots in six cities. It also involved a separate analysis, by academics at Manchester University, of a database of more than 2.5m riot-related tweets. The second phase – to be completed in 2012 –, will involve interviews with police, court officials and judges and a series of community-based debates about the riots," explains the Guardian.
The interviews were conducted by a team of 30 researchers selected through the Guardian's website and trained with funding from the Joseph Rowntree Foundation and the Open Society Foundation. Contact was made with 1,000 individuals who had been involved in the riots. The Ministry of Justice also granted access to 13 people convicted for their involvement in the disorders. The 270 interviewees who accepted to take part in the study were asked questions such as how they first heard about the riots, how they became involved, how they communicated, what they did, why they thought the riots had stopped and how they felt about their actions months after the events.
After collecting the material, a team of five research analysts recruited by the LSE began analysing the content of the transcripts in order to try to establish particular themes and their recurrence: a list of coding labels was produced and their relationships recorded and displayed on a thematic map document which allowed the research team to visualise the overall picture around the root causes of the riots.
Phase two of the research study is strongly focused on the community aspect of the riots and is aimed at bringing the individual experiences of community members, police officers and representatives of the judiciary into open interaction through open debates. This phase aims at understanding the direct aftermath of the riots and how the courts system coped with it. The same interviewing methodology will be now applied to judges and magistrates, police officers, court staff and prosecutors and defence lawyers.
Although the project is work in progress, the Guardian's 'Reading the Riots' website already features interesting insights. Data and visualisations are available to explore and share. An overview of the most captivating projects is provided below.
Interactive Timeline of the Riots
The timeline allows readers to navigate the unfolding of the events over time. It is linked to relevant articles published by the Guardian. Fires, riots, political intervention, court rulings and clean-ups are all displayed and cover a period spanning from August 4th to September 1st, 2011.
The map provides an immediate visualisation of the density of the events
The Riot Commute Map Animation
England's riots: Mapping the distance from home to offence
In explaining the methodology applied to the study the Guardian writes that:
"The data came from 1,100 individual's magistrates' courts records collected by the Guardian. For around 600 of them, postcodes for both home and offence location were available. The ITO World (the UK's top transport data mapping company) analysis is based on a smaller sample of 400 where both postcodes were detailed enough to allow exact mapping. There are some hefty caveats: we don't know for sure that those accused of rioting took these routes, or that they left from home. This is, in effect, a model for what might have happened. According to analysis by ITO world - based on the Guardian's database of riot-related court records - the average distance from home to where defendants were accused of a riot offence was just over two miles, or a half hour walk."
The maps and data used for the animation are available at the Guardian's Datablog.
How Rumors Spread on Twitter: An interactive analysis
The aim of this project was to show the "birth and death of rumors on Twitter" on the basis of 2.6 million tweets provided by the social network and somehow 'related to' the riots for containing some key hashtags. The Guardian describes how the method behind this first selection worked: "With help from an interdisciplinary team of researchers at the Universities of Manchester, St Andrews and Leicester, we distilled the overall corpus down to a series of subsets related to each rumour. We then undertook a more hands-on approach to find the tweets that best represented each story." The graphic part of the research was inspired by projects like Bloom's Fizz, which represents Tweets as circles grouped into larger circles.
"In our case, this grouping would place the items into clusters - each comprising a set of retweets for a given tweet. To make this work, we needed to find which tweets belong to each cluster. Again, our academic partners proved invaluable, providing a parametrized Levenshtein distance algorithm for finding all tweets within a certain "distance" from each other in textual terms," describes the Guardian article explaining the story behind the project.
Developers Martin Shuttleworth and Robin Beitra built an interactive timeline that would allow each rumour to be replayed like a video and Beitra also built alternative renderers for WebGL, HTML5 Canvas and Flash so that even older browsers would have access, thus granting access to the visualisation to the biggest possible audience. "Martin did some great work to make an interactive playhead that lets the rest of the system what time it is. Backbone.js proved very useful for keeping everything in sync. We added a graph of tweet volume over time to help people find the most interesting parts of the story. This is drawn in SVG or VML depending on browser capability," adds the article in the Guardian, which also defines the project as "one of the most ambitious pieces we have ever built, both in terms of data analysis and dynamic graphics."
The birth and death of rumors on Twitter during the riots