“Whatever the future holds for accuracy and journalism as a whole, it seems certain that the river of errors and corrections will continue to flow for decades to come.”
Jeff Jarvis (2007)

It could be so simple. The journalist works accurately, his texts are wirtten to the best of his knowledge and belief. And if there is something wrong, his colleagues and readers will point that out. The journalist thanks them for the hint, corrects the error and has made the world a little bit “correcter”.

Well. There are doubts about the accuracy in journalism. Texts are produced under time pressure. The fact-checkers are retired, their colleagues are busy and the readers indignant. The journalist is annoyed by the hint, he might correct his error and blames the news business. Although you can hardly find the type of journalist described in the first place, there is something you definitely come by regularly: errors.

It’s these mistakes Corrigo is focused on. Of course, there’s a high degree of professionalism in online journalism and you can find both extremes of the case scenarios described earlier.

Technological change has not only increased the need for quality control – but made it possible in the first place. The internet in general and social media in particular enables all citizens to participate in quality control.

Good News – Bad News

That’s the good news: There are millions of fact checkers out there and you always find somebody who knows it better.

The bad news is: The european research project MediaAcT has shown that many of the established tools of media accountability – including the press councils – fail to systematically involve citizens.

Without the public’s voice, these instruments are too quiet. Only as a common orchestra, supported by the mighty choir of the public, they can be heard.

The Corrigo Orchestra

There are already some pioneering projects like MediaBugs and NewsTrust, which share some of Corrigo’s ideas. While their services vary considerably in function and degree of professionalization, they have one thing in common: they do not reach many people.

And as long as Crowdsourced Media Accountability Services do not achieve critical mass, publishers and media professionals will not have the pressure to diminish the number of errors.

Still: Pressure alone would not lead to less errors and transparency. If the Corrigo orchestra wants to be heard it will need the right “tonality”. This is one of our key findings  after analysing error culture in journalism. Instead of “crowd versus Journalism” Corrigos slogan should be: “Journalism featuring Crowd”.

Corrigo is a Crowdsourced Media Accountability Service that helps its users to flag and correct factual errors, missing links and typos in online news sources. It comes as a browser add-on based on web annotation technology. This makes you see errors and corrections right in the context. To give you a better impression of our idea we produced this screencast:

Strengths

• Corrigo increases the visibility of errors in the direct context of the article.
• Corrigo accelerates the process of reporting and correcting errors and helps to document it.
• Corrigo separates error messages from comment threads and helps editors, to identify them easily.
• Corrigo lives and breathes in the browser.
• Corrigo stops the spreading of errors by marking them in correlated texts.
• Corrigo enables a comparison of the accuracy of online media.

Weaknesses

• New browser versions bring a need for constantly updating the required add-on with them.
• Web annotation systems are barely known and used.
• To sustain further development, communication and technical infrastructure, Corrigo is dependent on funding.
• Up to now: Corrigo is not working in Apps.

Opportunities

• Corrigo is able to establish a more open in dealing with errors and to increase the quality in online media.
• Corrigo has the potential to build a network of existing instruments of Media Accountability.
• Corrigo can be supplemented with additional function, for example, connections to services such as churnalism.com or tineye.com, web services, looking for the origins of texts and images to find unlabeled news releases and stock material.
• Journalists may use Corrigo pro-actively. With the help of Corrigo they could ask the crowd to verify a statement in an interview or a text.
• Via plugins Corrigo can be connected with content management systems, in order to enable an efficient management of errors and corrections within a few clicks.
• As a non-profit open-source project Corrigo can be internationalized and extended functionally.

Threads

• The existing culture of errors can complicate the introduction on the market.
• Despite all methods of quality control within the community: An abuse of technology for can never be excluded.
• Some legal issues are still unclear. Court decisions may change the legal framework to the disadvantage of Corrigo.

For us this is an interesting issue as we do not want to burden editors with even more work. Some of the editors and journalist we talked to feared that Corrigo would ultimately mean more work load – and we take this seriously. Still, we believe Corrigo could be a real enrichment in terms of collaboration between “the users” and the newsroom:

Good collaboration depends on good communication. Having examined the error culture in newsrooms it became clear that journalists have difficulties with admitting errors. That’s nothing special. Nobody likes to say: Sorry, my fault. Because we live in a culture of blame-and-shame. There’s a bit of Nelson in all of us as we love to point a finger on people who made a mistake.

The better way would be to blame the error, not the person who made it. If Corrigo was the place where readers pillory the writers the latter would not accept the new tool and nothing would change.

That’s why Corrigo’s motto is not journalists vs.crowd but it’s journalists ft. crowd.

As we mentioned earlier, “we take the technology of web annotation but limit its use to reporting errors in news articles.” We wish to focus on factual errors. They are “deviations from objective fact” and are therefore called “objective errors” (cf. Blankenburg, 1970:376). By concentrating on those objective errors we leave apart all subjective errors which are subject to numerous debates – which can be continued within the comment threads. This would enable the editor to access reported (factual) errors more quickly.

If someday Corrigo gets really really popular, we will probably have many “potential” errors reported. In order to keep the correction procedure efficient we provide a “traffic light system”. Users who have acquired a certain reputation within Corrigo will have the power to verify errors reported by users with a smaller reputation (e.g. new users). So if there were many errors a day the editor would be able to only concentrate on all red / verified errors.

Last but not least a short note on friday’s lecture. Oliver Reichenstein talked about several web projects his design agency Information Architects had realized over the past years. Among the relaunch of Zeit Online in 2009, which was notably interesting for me. Having worked in Zeit Online’s newsroom as an intern when the “switch was flipped”, it was great to have the designer’s perspective towards it. If you’re interested: Knight-Mozilla learning lab recorded the lecture, and I encourage you to watch it!

And that’s exactly what we think would be sufficient regarding Corrigo.

The technology Corrigo is based on is as old as the internet itself (cf. Peiwen, 2008: 2 ): web annotation. We chose web annotation because it brings some advantages which are crucial for Corrigo’s core functions:

• highlighting (good for pointing to the paragraph in question)
• annotation (e.g. sticky notes)
• independence from the annotated document

The two first points are important for the user experience because it will show annotations right in context at the article. The last point is important because it really gives the power to the people. We want to crowdsource media accountability and encourage free speech. That’s (only) possible if we do not rely on any implementation on the part of the publishers.

Another nice feature is the fact that the user does not have to pro-actively click on a button such as “Show me annotations for this article” because the browser add-on would automatically load any annotation when there’s data to that specific URL.

Funny enough, all this already exists. The probably best known service using web annotation is Diigo. But there are (or often: were) many other web annotation systems.

Also the idea of reporting “media bugs” is not new, but either depends on the pro-active clicking of a bookmarklet or on a (wordpress-)plugin installed by the publisher.

In our research we could see that there were many good projects – but none of them ever got to the critical mass. We think, this is why:

• most web annotation services are too powerful
  users were probably overstrained by the large feature set. Proof: check any “big” site like Google’s front page and enjoy the public Diigo annotations…
• fact-checking services like Mediabugs are a click too far away
  users have to actively check for new comments but are not pointed to the information upon clicking an article

So how could we do things better? Well, by tailoring and stoating these two smart approaches:

We take the technology of web annotation but limit its use to reporting errors in news articles.

What Zuckerberg later explained was that you cannot “create” a certain community because the community already exists. All you can do is to provide “smart tools” that help people to connect and communicate.

We often thought about this simple but important theory. Before starting to evolve the concept of Corrigo we did a big deal of research / market analysis. What we found were concepts that were never realised, venture capital driven start-ups that did not survive its seed funding or even projects that still exist but certainly could use more user interaction.

So in the end one thing became clear: Critical mass is, well, a critical subject. To us this result was not that surprising as we experienced, built and tested several (journalistic) projects that failed to reach a sufficient audience.

But regarding Corrigo, we see at least two “communities” that we have to think of. Besides the user who is supposed to mark and correct factual errors, there is also the undefined group of hackers that we hacks do count on.

One thing is clear: This group has to be the open source community as Corrigo has to be open source. In our thesis, we follow the idea that media accountability has to rely on the people and thus could not be the product of a privately held company. We think, this fits the spirit of the open source community.

Long before we applied for the Knight-Mozilla challenge we read through the Mozilla Manifesto because we wanted to understand how “they” feel about open source. Chris Heilmann also referred to this manifesto and described Mozilla as the “Red Cross of the Internet”.

I like this comparison in terms of openess, image and usefulness (not so much in terms of fundraising though :)). Still, the fact that the idea of Corrigo and the spirit of open source  fit quite well, this “does not guarantee that hordes of active developers will suddenly
volunteer their time” to our project.

For reaching the critical mass among the hackers we already know the wrong question: “How to build an open source community for Corrigo?”.

But fortunately there is a good “homework” to work on for the final of the Knight-Mozilla learning lab: a software product proposal.

Hacks usually fear to dive into the world of hackers. So it was good to have Christian Heilman, a Mozilla Developer Evangelist, focusing on HTML5 and John Resig, the creator and lead developer of the jQuery JavaScript library, in the boat. They pushed us into the ocean and showed us what’s under the surface of the water: the fascinating world of the open web.

John Resig gave us some interesting insights in the conception and realization of an open-source software project, which is first and foremost organizing communication and documentation in a smart way. Concerning Corrigo this is going to be a real challenge.

Corrigo is planned to be a tool to control quality in journalism. It’s our democratic conviction that such a tool mustn’t be in the hands of the state or the economy. Since the media as the “fourth force” control the government forces, the state in turn and in the sense of “checks and balances” may have no means of controlling the media. And as quality control of media is too important to be run by enterprises with economic interests it has to be run by the public, which is – going back to the philospher of the Enlightenment John Locke the supervisory of a balanced political system.

This means that while building and coding Corrigo as a non-profit and open-source project it’s important to find a balance between the liberty of the open-source community and the goal of creating a tool which empowers the public to take part in controlling the quality of the media without giving them any chance to abuse this power.

As John Resig made me think about the development of open-source software and let’s say some political or economical aspects of Corrigo, Christian Heilman inspired me to rethink the technical realization of Corrigo. Heilman told us what’s possible with today’s and tomorrow’s browsers. He’s a big fan of applications run by the browser because you don’t have to install software or to update a plugin. At the current state of conception we think that a browser add-on is the only way to realize what we have in mind.

Corrigo shall enable its users to flag and correct errors, missing links and typos in online news sources, so that other users see errors and corrections right in context (to get a better imagination of our intent we recommend to check our presentation on slideshare including some first mock-ups). We think that web annotation is the right technology for that. But is a browser add-on based on web annotation technology the only possibility? After having puzzled a playing music video in my browser we’re not sure anymore. There must be ways we haven’t thought about. So let’s go diving! Share your thoughts with us.

And that’s where Burt Herman steps into our virtual #MozNewsLab classroom. He is a journalist and entrepreneur, he co-founded Storify as well as Hacks/Hackers, an organization and network that brings together journalism and technology. Taking Storify as an example, Burt gave us advice on how to realize our ideas.

By the way: As I already had the chance to work with Storify, I have to say that it’s a smart tool for the kind of journalism, Jeff Jarvis once named process journalism. Due to the fact that “online, the story, the reporting, the knowledge are never done and never perfect” building a tool which helps journalists picking up tweets, blogposts, videos and photos, to rearrange and to comment on them is a great idea.

But as Jeff Jarvis’ lecture will come soon enough, let’s get back to Burt Herman’s advices. Let’s focus on these: Build a community, build a team and hey, just build it! So far our team consists of two hacks. Now it’s time to find some hackers and Burt Herman encouraged us to do so.

To pick up on Kersten’s thoughts on Aza’s lecture about prototyping, a maker space or a hackathon could be a good start.

See these many buttons up there? We probably did the worst thing you can do about buttons: We printed them on paper.

Of course we also put them into our presentation, being the essence of our thesis, which explains its functionality by telling a little case scenario. But eventually this is not satisfying anymore. We want to click these buttons!

Or like Aza Raskin stated in our first lecture on Monday: The value of an idea is zero unless it can be communicated. And a great way of communicating is to make people touch it. Aza illustrated (well, not he himself actually) his credo by providing a nice little cartoon which led him and other participants to the conclusion: Prototypes are like the tits of software design.

Well. As we are virgins in terms of software design this really is a challenge. During the whole course of writing our thesis we were kind of sad about not having the possibilty to easily code a rudimental prototype. It is nice to sit together brainstorming and dreaming about a certain idea that you believe in. But it is nothing you can really touch. And when you cannot touch, it is hard to get others involved or simply get some feedback about your theoretical construct. It is okay that the outcome of three months’ work is “just” some mock-ups as we did know that prototyping would have gone beyond the limits of our study.

But hey, study is over! Now is the time to make it real. With the motivation from our first week  at #MozNewsLab we are now thinking about how to get Corrigo clickable.

We will certainly talk to the few coders we already know. At the same time we like the idea of a “maker space” where people can start hammering on sofware ideas. It is our job to inspire participation now. So if you want to share your tool box – we are happy to provide the work bench.

One hundred sixty-two pages – that’s our final output of four years of online journalism at h_da University of Applied Sciences. And it’s also all about our visions and whishes that constitute Corrigo.

We already passed our oral exams and are now looking forward to our graduation (party!). By Friday you can call us Diplom-Online-Journalisten.

And there’s even more we can celebrate: We actually made it and can advance to the next stage within the Knight-Mozilla challenge “Beyond Comment Threads”!

We are really proud to be chosen to join 60 other participants attending a four-week learning lab. There will be two mandatory lectures (entirely online) each week – including speakers like John Resig and Jeff Jarvis. Awesome!