This week Gawker reported on a piece of investigative journalism in the New York Times which was estimated to have cost $400,000

“That $400,000 number comes straight fromTimes Magazine editor Gerald Marzorati estimating what it would’ve cost…had the Times paid for it in its entirety. But the bill for it was heavily footed by ProPublica, a independent non-profit “newsroom” doing investigative journalism.”

Journalists are used to dealing with inflated numbers and this certainly looks high. Other than the reporter working for 2 years on the story I’m not sure what other costs are included, but I’m open to suggestions.

Meanwhile, in June Monday Note estimated another piece of investigative journalism - this time in Sweden - at $50,000. This one took less time, but appeared to involve more people, and the figures are more useful to go through.

Both stories talk about the high cost of investigative journalism as some sort of defence: ”What’s going to happen to investigative journalism because of blogs?” asks Gawker.

But both start from a flawed assumption: that an organisation always needs large amounts of money to pursue investigative journalism.

Money, of course, is shorthand for everything it pays for: journalists’ time; overheads; expenses; experts; legal advice; and so on.

Go beyond the headline figure, then, and you can ask a different question: “How can we do investigative journalism cheaper?”

Or indeed: “Who can do investigative journalism cheaper?”

When that becomes the question, you think in a very different way. And that’s why the likes of Talking Points Memo have demonstrated there are ways to do investigative journalism much more efficiently.

But the key point for me is not just the cost, but the end result. As Gawker puts it:

“Just as important, however, are the number of readers who are going to click through to Fink’s story. Bringing into question a $400,000 story’s costs misses the larger point of how much the actual information contained within the story’s worth to readers, and who’s going to capitalize on it.”

And there’s the rub. Gawker points out that for all the money spent on that story, most news organisations will sum it up in a bullet point - and most readers will go no further than that. But when the likes of Talking Points Memo make an investigation a social act, engagement (and readership) increase accordingly. ]]> http://onlinejournalismblog.com/helpmeinvestigate/blog/how-much-does-journalism-really-cost/feed/ http://onlinejournalismblog.com/helpmeinvestigate/blog/critical-review-hmi-and-game-mechanics/ http://onlinejournalismblog.com/helpmeinvestigate/blog/critical-review-hmi-and-game-mechanics/#comments Thu, 27 Aug 2009 15:39:45 +0000 paulb http://onlinejournalismblog.com/helpmeinvestigate/blog/?p=124 In the second of this series of posts I want to look at how Help Me Investigate used game mechanics - how they worked in practice - and some of the plans for future development in this area.

Too small to be a game

When HMI first quietly went ‘live’ there were almost no game elements. We withheld from introducing ranking as there was no need for ranking when the community was very small. And the initial interface for challenges simply listed them, most recent top. The main game element was the concept of ‘challenges’.

As the site and its users grew, we changed the interface so that challenges started on the left hand side, then moved to the middle when accepted, and finally to the right column when complete (complete with big green border and tick). This made it easier to see at a glance what needed doing and what had been achieved, and also introduced a level of innate satisfaction in the task. Users, the idea went, might grow to like to feeling of moving those little blocks across the screen.

We then introduced a basic ranking table to the investigations page based on a measure of ‘activity’  of users.

Phase 2

Essentially most of the game mechanics planned for the site were not core to its function, and so we have held them back for Phase 2. These include:

• More sophisticated ranking system with associated permissions, e.g. someone with a lot of authority may be allowed to perform a ‘verify’ challenge
• Possible differentiation of challenges based on difficulty with points awarded accordingly
• Possibility of points being able to ‘unlock’ things either on the site (e.g. personalisation, bonus features) or on other sites.
• ‘Badges’ awarded for completing different types of challenges and displayed on a profile (e.g. FOI, law, etc.) Acquiring these may lead to you being ‘recommended’, for example.
• You can ‘buy’ (with points) ‘Hints’ on challenges to help you.

Listed like this they make the site sound much more frivolous than the reality. I’m not saying we will necessarily do any of this, but these are certainly ideas we may try to make investigating innately satisfying and rewarding. We’ll see how they work, and if they don’t, we’ll stop.

What sort of community?

One of the most exciting points of HMI’s development was when we started to see a genuine community emerge with a shared interest. In this case, that shared interest was in data, and emerged in large part because of a couple of particularly active users with that interest who fed off each other’s activity.

What was curious about this was how it raised the question of ‘What sort of community is this site designed for?’

Now, using agile development meant that, while we were designing for particular types of user, we also adapted to react to the real users who used it. But would whatever community first dominated skew the site away from other types of user?

At least one user commented that the site was very ‘FOI-heavy’ and risked excluding users interested in different types of investigations, or who saw FOI as too ‘highbrow’ for them.

So the question is: do we aim to support other communities within the site to ensure a broader range of investigations? Or do we focus on that niche of data/FOI-focused investigations?

We’ve partly played with this by bringing in a second support journalist with a consumer background - Colin Meek - to complement the FOI/data strengths of Heather Brooke.

At the same time we’re actively involving users with other investigations, and as more users have come onto the site these have naturally broadened into consumer, property and law. I think the initial answer is that the community is not large enough yet to be dominated by a particular strand, or for us to take that decision on whether to focus on one niche.

Not social enough

The main disappointment of these first few weeks has been the lack of users getting involved across investigations. Users tend to stick to their own investigation and the idea of ‘helping another so they help you’ has not been established.

This may not be so unusual. I’m waiting for some analytics from the site, but I suspect there may be a power law distribution at work here - often seen on the internet - of a few people being involved in lots of investigations, most being involved in one, and a steep upward curve between.

The community manager, Nick and myself have invited people to investigations where we can see a shared interest or location; and Stef has developed a ‘related investigations’ section at the bottom of every investigation, as well as a ‘recommended investigations’ section on the homepage based on a user’s tags.

There are also a lot of technical features we have held back from developing which would encourage more cross-investigation activity, including being able to invite other users to an investigation (currently you have to know their email address), and being able to invite users to a particular challenge you think they’d be good at. I also wonder whether distinguishing between ‘proposers’/'instigators’ (those who start an investigation) and ‘helpers’ (those who join) - through profile ‘badges’ for example (more on badges in a future post) - may be helpful socially, creating a social pressure not to merely start an investigation but help on others?

I have also not seen as much evidence of users inviting others as I would have expected. Partly this has been through a fault in site design - the ‘Invite people’ button was not in a prominent enough position until a few weeks ago and it’s currently not possible for users to send a general invite to the site. A third issue is that some invite emails are being filtered into Junk Mail folders. This is simply a budget issue, and in a fully funded development phase we would address this.

Technical issues aside, I don’t feel we’ve done enough culturally to encourage invites and promoting investigations. We need to set more ‘invite others’ challenges; we need to be scouring Facebook groups, forums and blogs for more potential investigators; and we need to be helping users to campaign for their investigations to raise awareness. On all fronts I suspect the ‘invite-only’ status of the site has held us back mentally and technically. That may not be a bad thing as we focus on building the culture of the site but at some point we need to be brave in deciding it is strong enough to take on allcomers.

Social = engaged

One of the concepts behind the site was that if you had a social connection with an investigation you were more likely to become engaged in it. This has certainly been born out in practice. The first time I experienced this was when I joined an investigation by Donato Esposito into two bus companies running buses on the same route. Apart from my professional role on the site I knew Donato as a friend and that dynamic led me to learn all sorts of information - on bus deregulation - I would otherwise have never explored.

Likewise, when I started an investigation into hospital parking charges it was not done with particular enthusiasm. But when another user dug out some useful information from me I felt a social pressure to ‘pay back’ that effort with my own. I became more engaged in the investigation, more active, and have dug out increasingly interesting information as a result.

That’s been incredibly heartening. On the other side, however, is the tragedy of the commons - when people feel their efforts are being exploited elsewhere. Journalists are notable for taking more from the site than they put in, and this is a cultural issue that we are addressing on an ongoing basis through communication and workshops. At the very least journalists have added official reaction and a formed narrative to the stories that they have taken from the site; but that still generally requires someone to chase them to add it to the site. One journalist has contributed enormously to the site; another was particularly helpful in obtaining raw data on a story that was doing the rounds of the local newsrooms. I have met with dozens of journalists to explain the site, but it will take hands-on workshops and institutional recognition and support to truly change this. Thankfully we are already on the path to both of these with more than one news organisation.

The original idea behind HMI was inspired by investigations like these. Imagine, I thought, if there was a bespoke tool designed to allow people to collaborate on an investigation, rather than these crude tools that were built for another purpose?

But now that HMI was actually here, being used - apparently successfully - I needed to ask myself the hard question: we’ve imagined, now how does it work in practice?

Ultimately the answer came when I was creating a blog supporting one of my investigations - into hospital parking charges. At the end of every post I invite people to help and I realise that there were a number of problems:

• Firstly, it requires a lot of effort over a long period of time from a blogger to get the necessary number of readers of the blog to have just one who is able and willing to contribute. Bloggers with an existing audience have an advantage, and of course you could piggy-back on such a blogger’s popularity, but that option isn’t open or obvious to most people.
• Secondly, when a potential collaborator reads your blog, it is not immediately clear how they can contribute. Even if I, as a blogger, created the equivalent of challenges, the blog interface does not make it possible to display those in a way that is easy for users to navigate.
• Related to this is the ‘begging’ issue. After 3 posts that end with a request for help, my blog already feels kinda needy, almost spammy. And I’ve noticed that other people who blog about something they want to investigate rarely ask for help, but rather tell the story.
• Likewise, the nature of problem #1 means you need to provide regular content, which would push ‘challenges’ down. This also produces a new barrier to entry: if you are not a natural producer of content, or write in an incoherent rambling way that obscures the clarity of what you’re trying to achieve (as many angry people do), then you are at a disadvantage.
• Next problem: even if someone agreed to help there is no easy way to support that or monitor who is really doing what and who has forgotten.
• Finally there is the question of how we get the ear of the powerful to force change.

There are similar problems for forums and Facebook groups. Investigations that have successfully used these tools have overcome all of these problems (the Stirrer forums are one of the most successful examples in that they have a professional advocate in site owner and journalist Adrian Goldberg, who is willing to take on causes and use his experience and knowledge to drive them forward. The problem is that he is only one person) - the question is how many have stumbled at one of those hurdles?

When I looked at Help Me Investigate in the context of these problems I could see that, in a nutshell, it was lowering the barrier to entry for people who wanted to investigate something:

• A community is already in place who have effectively stated that they are engaged in issues of public interest. You still need to build an audience and activity around an investigation, and blogs/forums/Facebook play a part here, but there is support for that based on the experience of other users recorded in the site KnowledgeBase.
• The interface makes it easy for potential contributors to see how they can contribute to an investigation - as well as suggest routes of enquiry for others.
• The ‘Help’ is explicit in the site and implicit in the investigation. No pleading required; no apologies. The clue is in the name.
• Content doesn’t have to be your problem. HMI manages the results and process - content is just one challenge for anyone to produce. That said, currently the site needs more people doing this.
• You can monitor who is doing which part(s) of the investigation, and there are plans to build in reminders, expiry dates and other features.
• The involvement of journalists and news organisations in the site means they can press for official reaction and action. (There are plans for users to specify whether they would not prefer the involvement of mainstream media).

The core function, of course, is to make it very easy for someone to make a very small contribution to an investigation without needing to be as engaged as they would need to be in a ‘traditional’ blog-led investigation.

It is also to provide the social stimulus to encourage someone to act on that question they have, or injustice they’ve experienced. Rather than feel they are blogging into a vacuum, the simple act of being set a challenge, receiving an email from the support journalist, or having someone sign up to your investigation may be enough to incite them to send that letter or make that phonecall. And from there to help others.

And of course it is to collate and connect all of that experience to provide a platform for others to make a difference without having to start from scratch every time.

Of course the successes of HMI in just a short time proved it worked, but it would be unprofessional to accept those successes at face value. I’m naturally distrusting of good news stories. Help Me Investigate is at this stage just a rough proof of concept, and has its flaws. Its success has surprised me - it has far exceeded the basic expectations we had, and that is before we implement any of the many features we planned and put aside for ‘Phase 2′. The key thing is: it works, it serves a different purpose to the blunt tools that already exist while not seeking to replace them, and it has incredible potential to empower and engage people.

Now we’ve got that sorted, let’s keep going.

“Anyone want to send their first Freedom of Information request? It’s already written for you. Invite to @helpmeinvestig8 thrown in.”

The rest of the conversation is captured below in a screengrab, but here’s my summary of the positions:

Steve: “each FOI potentially takes hundreds of pounds to answer. You have to weight up need to know with that … this “roll up roll up let’s file an FOI” approach just doesn’t seem responsible to me.”

Me: “In this case, the FOI was going to be sent and I thought it was a good way to show someone how to do it. Call it distance learning.”

Jon: “I’m pretty well educated and wanted to make an FOI and I found it a little daunting, luckily I got support from people like @rasga”

The debate continues, but I thought a blog post and comments might be a good place to keep a record, rather than it disappearing into the Twitter cloud…

Steve Jackson (ourman) on Twitter via kwout

Paul Bradshaw (paulbradshaw) on Twitter via kwout

jonhickman (jonhickman) on Twitter via kwout

You can find the widget listed on Widgetbox here. Click on any of the boxes to the right which will help you embed the widget on Facebook, Blogger, Wordpress, MySpace, an iGoogle page, and various other services.

In future I’ll put together widgets for the latest reports, how-tos, and other parts of the site.

Q: “Journalism, especially investigative journalism, is labour-intensive and expensive. citizen journalists cannot expend the time required to uncover what centres of power are really up to. “

A: “…In just a couple of weeks we’ve uncovered a £2.2m overspend by Birmingham City Council, the most ticketed places and most over zealous parking inspector in Birmingham, and the fact that BCC use surveillance powers 6 times more than they should be based on population. So don’t tell me what is “required to uncover what centres of power are really up to” without any evidence.

What I’m trying to do is find out just what *is* required and possible - and I’m actually quite surprised what we’ve managed to do with very few people and little resources (this is a prototype stage).

Also, this is not citizen journalism - it is a collaboration between professional journalists and non-journalists with fantastic expertise and passion. For example, most journalists aren’t able to analyse data in the way that some HMI users have.

I don’t think it’s the only way to do journalism, and I think there are forms of investigative journalism it couldn’t do, but it is another way of pursuing stories and digging up evidence that takes advantage of web technologies rather than fighting against them.

Finally, this isn’t some bloody competition, but an attempt to help journalism and journalists in practical terms rather than blindly hoping for some magic bullet that will keep things as they are.

It’s also an attempt to engage people in local and national issues and give them the knowledge to challenge power. That’s the sort of thing I’m passionate about in journalism.