Metamedia

Exciting new projects

I’m stopping mid-travel on the way to spend a long weekend making a newspaper in a field, because an exciting new thing happened today.

I’m going to be teaming up with some of the guys at Neon Tribe – a local web dev firm who make awesomely exciting shiny Internet things with open data – to learn more data skills. I’m hoping to collaborate to bring a few of my long-brewing ideas to fruition, make a few shiny things together, and ultimately learn and improve the skills I need to make shiny things all by myself.

It’s not clear yet how much time I’ll get to spend doing this – most if not all will have to be outside my standard working hours as there are heavy constraints on resources in the newsroom right now – but even if it’s just an hour a week being a hack in a roomful of hackers, absorbing, learning and imparting with a few projects to focus on, it’s going to be a very exciting and – I hope – productive time.

Three-tier journalism

There are three tiers of journalism in the UK at the moment – national, regional and hyperlocal – but in all the discussion and excitement over open data, the voices of journalists working at the coal-face in the middle tier tend to be absent. That’s a shame, because regional news offers some fascinating and unique challenges for data journalism and computer assisted reporting.

At one end of the scale there’s national journalism, which covers big issues affecting all regions of the country or stories of national interest. In most media national journalism tends to be biased towards the south in general and London in particular, and in newspaper terms there’s a partisan/issues bias too, along with a clear character.

Then at the other end of the scale there’s hyperlocal journalism, geared around my street, my postcode, my community. These are organisations tackling incredibly specific situations, interested in minutiae and detail, as well as the impact of wider stories on the communities in question. It’s all about applying the national news to a very specific set of circumstances.

Somewhere in between, on a sliding scale depending on the size of the news organisation, is regional journalism. At the moment that’s where I fit in – at the city- and county-wide level depending on which paper I’m writing for. The stories I follow up are a mix of both – national stories with an impact on the communities I write for, and street-level stories with wider implications. We also cover wide regional stories with an impact on a substantial proportion of our readers – council stories, crime cases, the sorts of stories which nationals would not cover at all while hyperlocals would cover only the relevant parts.

After a conversation with the BBC’s Martin Rosenbaum at Hacks and Hackers, I started to understand that regional journalism has a particular set of needs and problems when it comes to data journalism. National news needs big picture data from which it can draw big trends. Government ata that groups England into its nine official regions works fine for broad sweeps; data that breaks down by city or county works well too. Hyperlocal news needs small details – court lists, crime reports, enormous amounts of council information – and it’s possible to not only extract but report and contextualise the details.

Regional news needs both, but in different ways. It needs those stories that the nationals wouldn’t cover and the hyperlocals would cover only part of. Data about the East of England is too vague for a paper that focuses primarily on 1/6 of the counties in the region; information from Breckland District Council is not universal enough when there are at least 13 other county and district councils in the paper’s patch. Government statistics by region need paragraphs attached looking at the vagaries of the statistics and how Cambridge skews everything a certain way. District council data has to be broadened out. Everything needs context.

The great thing about that? There are unending opportunities for good data journalism in regional news – opportunities to combine new technology and open data to produce something that’s relevant and useful to as many individuals as possible. The question is how we exploit them. I believe that we start by freeing up interested journalists to do data work beyond simply plotting their stories on a map, taking on stories that impact people on a regional level.

How do school catchment areas affect house prices? Since the county council decided to turn the lights off at midnight on certain streets, has there been an increase in crime? How have mental health service closures hit NHS waiting lists in the region? We should be using open data and freely available tools to do good regional journalism and helping people to find out.

Hacks & Hackers London – a few thoughts

Last night I headed to London for the first Hacks and Hackers meetup in the country. It sprouted out of the Ruby in the Pub meetings and has been organised by Joanna Geary and colleagues at the Times as a meetup for journalists interested in coding and geeks interested in journalism.

It was very well attended – more than 50 people by my count – and full of interesting conversations to be had about what people are currently doing and what they want to move into. I hope it continues and develops into a more structured, regular meet – and that I don’t have to get up at 5.30am to do an early shift every time I want to get there.

A few things I took away:

• It’s obvious, but it bears repeating: ideas are easy, execution is hard. We need skills as well as inspiration to build what we want to build.
• The corollary: collaboration is key. We don’t all need to know how to do everything – what’s important is building working relationships with people who can make up for our own shortcomings.
• Regional data journalism is a massively different prospect from hyperlocal or national. This is something I need to parse out, probably in its own post, but an important realisation – regional data and its implications and interpretations have their own joys and challenges.
• There are not many regional data journalists, especially in print-first organisations. To my knowledge I was the only person there who works on a regional news organisation, never mind a newspaper. I suspect there are plenty of reasons for this – time, inclination and understanding being just the most obvious ones. If we want to include interested regional print people in these conversations, we might have to take the mountain to Mohammed.
• Practical beats theoretical. If you want to persuade someone that data journalism is important or relevant, creating something is far more persuasive than explaining it. Don’t preach it, do it.
• Start simple. Everyone I spoke to who’s played a part as a tech/social media/digital evangelist in their organisation has said the same sorts of things – start with what’s possible right now, and work upwards from there.
• Not everyone with something to say is on Twitter or participating on blogs. Sometimes if you have a wide circle of social contacts it’s easy to miss voices from outside that circle – and often those voices have something tremendously valuable to contribute.
• People are amazing. Everyone who was at the meetup last night had some incredible ideas and projects on the go. Everyone had something to offer, something to teach and something to learn. At the risk of sounding completely hokey, it gives me hope.

And here are just a few of the ideas that I heard kicked around for what the group could do in the future.

• Hour-long lessons in all sorts of subjects – Ruby and other languages, Freedom of Information requests, story construction, search engine optimisation, data cleansing, social media, and so on.
• Talks or discussions led by people with practical experience covering topics that hit the industry at the moment – monetising online, social media policy (or lack of it), the fallout from the Wikileaks disclosures, for instance.
• A swap shop for people looking for help with projects or for learning mentors – almost a lonely hearts system for hacks seeking hackers and vice versa.
• Practical demonstrations and talks from people with proven experience.

Whatever direction it goes in, I’m glad to have been involved and excited to see what happens next.

The NCE News Practice exam: resources

I discovered on Wednesday that I’ve passed my NCE exams – and did particularly well on the News Practice exam, winning the Ted Bottomley award. (I love the name. Love it. Probably too much.)

The examiner [pdf] was very, very nice about my paper, saying:

A textbook example of how to tackle the Newspaper Practice paper. A comprehensive law
answer citing relevant cases and law, followed by practice answers that clearly demonstrate
the candidate’s imagination and ability. It is clear from this paper that this candidate is
already putting into practice the skills that the Newspaper Practice paper looks for. One of
the highest Newspaper Practice scores in recent years. A very impressive performance.

I’ve been trying to find the paper I wrote so I could work out what on earth I did right, but so far haven’t managed to unearth it. I’m pretty sure I arrived home and thrust it as far out of sight as possible along with the other papers.

But from what I remember, a staggering amount of what I wrote for the second half of the paper was about the internet. Specific, useful, relevant ideas about how to use it to move stories on, to facilitate comments and let the community take control of the conversation. I talked about topic pages, context as an integral part of news reporting, data journalism in many forms, visualisations, mashups, maps, timelines, social media, FOIs, online reportage in all sorts of guises and the importance of the hyperlink.

Anyone revising for News Practice exams – my best advice is read the links, think about how you can apply the theory to the practical, and good luck. Oh, and know your McNae’s. Nothing can beat that.

Braindump: just add points

Interesting presentation by Sebastian Deterding looking at what user experience designers can learn from game design.

Although news orgs face very different challenges from UX designers, the basic messages about shallow vs deep engagement, using multiple interacting points/currencies and measuring achievement, effort and attainment in a meaningful way are very relevant. Take a look:

It’s interesting to look at the Huffington Post’s community moderation badges in terms of this presentation. My gut instinct is that they fall, along with Foursquare, into a category of too simplistic game-like systems (“Just Add Points”) that don’t actually tap into the power and fun of learning that is one of the fundamental building blocks of good game design.

It’s also worth checking out this post on rescuing princesses at the Lost Garden. If you click through to the slides (PDF) there’s a thoughtful discussion of the differences between app and game design, and a very useful breakdown of STARS atoms – essentially, small chunks that introduce players/users to new skills, let them discover how to use them, and ensure they have mastered them.

Between them, these two posts and the thoughts behind them make a mockery of the idea of game mechanics as simple point systems you can pop atop pre-designed apps or comment systems or whatever it is you’re already doing. You have to design with exploratory learning in mind, with a learning curve that doesn’t flatten out horizontally or vertically and with end goals and nested goals to maintain engagement.

I wonder how the Guardian’s crowdsourced investigation into MPs’ expenses would have gone if they’d added this sort of rich game-led design? As well as giving long-term and short-term goals/rewards (like Twitter translator levels, perhaps) with status bars to show progress, perhaps they could have rewarded people who found something of real import with a status bump, or added exploratory learning elements by advancing users towards the goal of signing off on things other people had flagged as interesting. Or teaching basic maths, or collating data into a wiki-style “what does my MP spend” database, or encouraging/letting users learn to create their own visualisations of the data. Hard to say how well or whether that would have worked, but it’s easy to see wider possibilities in projects like that.

/end braindump

JEEcamp thoughts on data journalism

I spent some time talking to Martin Belam (@currybet) about data journalism and the importance (or otherwise) of journalists learning to code.

He said, as he’s said before, that it’s more important for journalists to know whether something is or isn’t possible than for us to necessarily be able to do it ourselves.

And for working journalists whose day to day job doesn’t carry a coding requirement already – and particularly those of us who are lucky enough to be in a workplace where there are developers or programmers who can take our ideas and make them flesh (ie. not me), he’s almost certainly right.

Those skills are becoming more and more important. With the birth of data.gov.uk and the increasingly open approach to information that the new coalition government is likely to take, sifting and analysing data to find the stories is going to be a vital skill for a lot of journalists.

We need to know our way around a spreadsheet. We need to be able to spot patterns in data and understand not only what they mean but also how we can use them to reveal stories that are not only relevant but useful.

We need to know where our skills can get us. We need to know our capabilities and our limits – and, crucially, we must be aware of what we don’t know. That’s not just knowing that there are holes in our knowledge, but knowing the shape of those holes so that we can try to get our problems a little closer to a solution.

Journalism is about asking the right questions. We research stories before we interview subjects so that we can ask pertinent questions whose answers will illuminate the subject. We need to be able to do the same thing with our data – we need to know what questions to ask and how, so that even if we can’t make the tools ourselves we can hand over the task to someone else without asking the impossible or wasting their time.

But most of the time, certainly for journalists on regional papers and I would wager for many in other areas, those people who know how to make the tools just don’t exist. I have friends who code, but I can’t ask them for a favour every time I want to create a news app, or diff two versions of a stack of documents, or visualise a complex dataset, or tell the story of 100 people’s losses from an investment fund going bust in a way that conveys both the scale and the humanity of the problem.

Regional journalists work on hundreds of stories that could be made vastly easier or more beautiful or more accessible through a touch of computer work (spreadsheets, maps, things that aren’t quite coding but sort of almost are and look like it to the untrained eye). A few of us can create those additions; the rest just write the story, and our papers and websites are poorer for it.

We work on a few stories – and the number is increasing – that are perfect for news apps, web coding, multimedia packages or other more complex solutions that very, very few of us can create. But no one else will do it for us.

On top of that many of us struggle with inflexible content management systems that penalise or make it literally impossible to display data-driven work online. Faced with that problem, some budding computer-assisted-reporters give up before they’ve even started.

So I’m not going to stop learning Python. It’s not a complete solution to the problem – for that we need real, systemic change so that the businesses we work for all value data work, understand its increasing relevance, reflect on current practice and support training journalists to do an evolving job.

But for me, it means that in the future I might be able to create better stories, automate processes within series or campaigns or multiple follow-up stories, make my job easier and make a better experience for the reader all at the same time.

At least, until we all have newsroom developers.

Emergent thoughts on emergent stories

After an interesting conversation with @harryharrold and @MrRickWaghorn yesterday, I’ve been mulling a few thoughts on emergent stories and how the social side of the web could make it possible to curate and (to some extent) formalise them.
Continue reading ‘Emergent thoughts on emergent stories’

Quick hit: data visualisations

Quick hit for the weekend: if you’re bored, I’d recommend checking out Smashing Magazine’s modern approaches to data visualisation.

They’re useful, quirky, fascinating, often beautiful, always unique, and seeing all these incredible projects together in one place is very exciting. And colourful. Definitely colourful.

If you’re still bored, take a look at Infosthetics. I guarantee you’ll be there for weeks.