 Excellent. Thank you very much. So, Jonas has done something of my job already because I'm here to talk to you this afternoon a little bit more about data visualization approaches. So, as you'll see, one of the main arguments is that there's been so much of a focus on the big data side and not enough of a discussion on the technologies and how those have evolved to visualize and represent and interrogate that data. So, it's a little bit difficult pitching to a very diverse audience here but what I've tried to do is organize a presentation that looks a little bit of an introduction to data visualization, some of the context and drivers for where this is coming from, some of the challenges we face in producing such visualizations, which some of you might have more experience of as well and would be happy to discuss in more detail later. And then to end with a few examples based on a competition that I've been running for the last two years with different think tanks and research institutes around the world and how they actually have been using data visualizations in their work. So, quickly an introduction to data visualization. I guess the first point is actually, are we calling it the right thing? Is data visualization what we're really talking about? And I'm sure you all know what this says, no? Well, wait, yes, some of you, you know, you're the well, it's a binary and maybe that gets into ASCII codes and ultimately we're getting into something that we can all recognize, which is hello. And that gets to something of the difference between the raw data and information and data is the lower level of abstraction from which information can be derived. I don't know if you know the French term for data, that points to the Latin root, data literally means that which is given and that is all that data is. But ultimately what we're trying to convey with data visualizations is information, something with meaning that is immediately legible and intelligible to the receivers of that visualization. So what are data or I suppose information visualizations? It will be no surprise to you that it's a visual way of conveying information, often quantitative in nature, although not always and would be great to hear if you have qualitative examples in an accurate and compelling format. One of the things that's really useful about information visualizations is that they have a great ability to make relationships between different data points more apparent. So that can be done by physically clustering items, putting them in a similar location. It can be done by using color or it can be done really importantly by putting information in scale. So you see a lot of things from I don't know if you know information is beautiful. One of the great ones that they do is budgets for different things and how much does the Iraq war actually cost. When you see that in scale compared to what you spend on economics or sorry on education or something, it really helps you to understand and comprehend that when you start getting into really big numbers like billions and trillions and petabytes, which we don't even know what that means, can't really visualize that or understand it. An important thing is that data visualization these days are kind of two different approaches to it. One is more on the static side and one is more interactive data visualizations that often draw from the live data that we just saw a great example of. And I think that that is an important distinction because they empower different types of activity. Oftentimes the static visualizations are designed to be more of a communication tool with a clear set of messaging. You might have heard the term infographic that are designed to convey a story or make a point. Whereas some of the more interactive visualizations can do that but are also designed to allow users to explore data and are useful in terms of gaining insights from the information that we have. And I think that it's important to put this in a little bit of context. Jonas has already hopefully done some of that. But the first is to say that it's not that the reason that we're here talking about this today is because data visualization is some massively new revolutionary concept. There are lots of examples I could have chosen. What I've gone here is the Coxcombe diagram from Florence Nightingale. This was used in the Crimean War to show that actually most of the deaths had nothing to do with war, which are all of these sort of injuries and deaths there, but actually much more to do with infections. And this is what helped prompt modern approaches to nursing and really did have a lot of impact. So it's not that data visualization is something new. But I would argue that there are three main elements of why we're talking about this a lot today. And one of them is big data. One of them is changes in the way that we access information. And some of it is the new technologies for producing these data visualizations. Now, as we've seen, we've just had a whole conversation about big data. There tends to be a lot of focus on there and not enough on these other two. So I will actually go ahead and focus just on these last two, starting on changes in the way that we access information. This is particularly important from a research institute's point of view or a think tank in that people aren't necessarily going to download PDFs of your report. People are increasingly using smartphones and tablets to access information. That means they might be doing it on the go. That means they might be doing it while listening to a presentation, at least one person. And so there's a different way in terms of how that is useful and how we interact with that. There's also disruption in terms of the media sources and outlets that we go to. So there's been a big explosion of data journalism. And you have websites like vox.com from Ezra Klein in the U.S. or Nate Silver's, what is that, 538, or the upshot from the New York Times that are all into data journalism and might soon be taking away our jobs from research institutes and think tanks. But also, I think that there's an important thing in terms of where people are going for information. There's been already a lot of talk about Google today, but actually I don't know if you've tried Googling recently and feel free to do so. Something like percent of people who have mobile telephones in a country that automatically goes into World Bank data sets and gives you a graph right within your Google search. So no longer are we bringing people to us, but rather we're going to where people are already accessing information in area places like Google. And I think that there's also with a lot of these a greater expectations towards interactivity. So you see children today that are growing up with iPads, going up to Windows and going like this, trying to zoom in. But that will only become more and more prominent as that generation grows up. And we also know that people really aren't downloading PDFs. I don't know if any of you saw this recent World Bank report, I guess it's from about a year ago, basically showing that about one-third of World Bank PDFs had never been downloaded. And of the rest, of those two-thirds, about half of them had been downloaded I think once or twice. I imagine that's the author checking to make sure that they're up there. I'm not quite sure. But I think that this focus on producing the report, producing the PDF is something that we really have to be questioning and interrogating right now for those of us working in the research community. So that's a little bit about how information access is changing. But also there's a big focus, a lot of changes in the way that we can now visualize data, and a lot of them are free or low cost, although some of the more powerful ones certainly have lots of costs. For example, we have things like Google Fusion Tables, Tableau you might have heard of and Tableau Public, CartoDB, Infragram, Pictochart, I mean the list really goes on. And that's why one of the things that I've done as part of the on-think tanks data visualization competition is not only create how-to documents in terms and videos on how to use some of those, but also really importantly reviews of those various technologies to see where they really work and where they fall short, how much they cost, all of that sort of thing. So I encourage you to check that out. But actually we've come a really long way in our ability to represent that data. But that doesn't mean that there are no challenges in production of data visualization. And I'm framing this in two slides really, and one of them is about the different skill mix that is required to make an effective data visualization. It requires everything from good research skills, and that is everything about ability to properly manipulate data. We were already talking about statistics, but research could be much more broad. It could be about social science methodologies. It could be about knowing the difference or something about more the context in which the data are situated. In terms of technology, we also get a big focus on technology that might be about knowing how to use these different programs and to use them effectively. So knowing the difference between an ordinal variable and a categorical one and things like that and being able to have actual coding techniques so that we can do that. Although a lot of the technologies mean that you don't have to be a coder anymore. We can talk about some of these different technologies because what I find particularly interesting is that some of them have been developed from a design point of view, some of them more from an academic research point of view, some of them from a really coding point of view. So they all kind of are better at different areas there. But we can't forget these other two really important elements of it. One of them is communication. That's about refining messages, editing. It's about really translating that process of finding data to conveying information. And there's design, which there's so many important elements of design from color schemes. People tend to forget things like not using red and green for simple colorblind reasons or things like that. So there's lots to be keeping in mind when we're making a data visualization and it requires a mix of skills that are difficult to find in any one person. So I think one of the important points on this is perhaps as an organization, it would be really useful to be thinking about how do we create effective teams that work together that bring the skill mix to the table. And this is not mine, but this is also a really important one from David McCandless. Again, I've already referenced information is beautiful. He says that a good visualization contains four elements to it. One is that it should be interesting, so it should be showing something new. A lot of times you get things like population heat maps that show it's just the population. So that's not actually anything interesting or insightful or new. It should be functional. There should be a purpose behind it. The form gets into that design element. And I think one of the really important parts of it is certainly the integrity, that you're representing data accurately and in a way that still conveys your message. So I have examples of those. I'm not going to go into them now, but it's something that we could certainly discuss later in terms of those four different points. So the last thing that I wanted to do was show a little bit of examples that I've collected from around the world from different think tanks and organizations. And like I said, this comes from the on think tanks data visualization competition, which we've run for two years now, where we're hoping to inspire, strengthen capacity, and encourage people to experiment with producing these types of data visualizations. In the most recent example, we had 32 visualizations from 31 different think tanks and in 19 countries around the world. So that's where I'm going to be pulling some of these different examples from. I forgot to bring the booklet up, but I do actually have a compilation of this year's entrance with some of the how-to's and everything if you're also interested in that. So I hesitate to call this a typology, but let's just say that here's different examples of things that you could do with data visualization. So one of them that we saw quite frequently was making government information or inaccessible information more accessible to a wider public. So one of the winners, not from this most recent year, but from the first year that we ran the competition was Scopia Rasta, which is all about town planning and Scopia Macedonia. And I don't think I have time to run the video here, but basically it's taken detailed city development plans and rendered them in 3D so that people can see how the city has changed. And really importantly, one of the things that becomes quite stark when you actually see this visualization is it takes not just what has already happened, it also takes a forward city plan and it just colors the green space green and it shows how quickly the green space within Scopia is actually shrinking. So I think that it becomes a really powerful message, although it's designed as a way that people can kind of play around and see these different things, but it's one of the things that is a really powerful takeaway from that. I should say, in case you're worried that this takes a massive amount of time and skills and everything to be impactful, this is an example from the Neffield Trust, which was one of my main clients, UK health think tank. All we did was within probably a couple of hours of the autumn statement last year, which is the UK's budget basically. They were making some key announcements on health spending and we really quickly put together the Sanky diagram with a little bit of analysis and we literally only put it out as a tweet and just on the back of that, the office of the shadow secretary of state for health getting in touch saying, can we have that? Can you explain this to us? Can you come talk with us? So it doesn't have to be sometimes these big massive years of analysis to be impactful. It can be something that, if you can just present something, if you can find that niche, can work really well. Another thing that people are doing quite frequently, and I'm sure you've seen examples of this, is using data visualization to communicate their own and really other research. So potentially mashing up different data sources. For example, this one was one of the winners from this year, which was called Don't Limit Her Possibilities, which looks at the country of Georgia, and does this big infographic comparing STEM achievement between boys and girls and how it starts to, although it's quite similar when they're young, it starts to diverge, and then they've married that with the data sets on the economics and skills gaps within Georgia, in particular in saying we need more STEM people, and you're discouraging that, and a lot of the messaging going back towards parents there. So that's one way of doing it. Another way that I've seen it, I've used the word splainers, that's just short for explainer. Some of you might know mansplaining, I don't know, that's where that word comes from. This one is a great example from International Plan Parenthood Foundation, where actually what they've done is they've taken their full in-depth reports and they've just put it into a data visualization, and what's really great about this data visualization is that it actually starts with, you press play, and there's effectively an audio tour of the information that is combined, and then once the audio tour is done, you can then start exploring and mixing and matching your own way. So there are different things that you can do with that. And then another one of my personal ones, and this might be an interesting thing that one could do with climate science actually, was tracking discourse. And so Ethos Public Policy Lab in Mexico created a visualization called Mexico en 140, which was Mexico in 140 characters, where basically they were taking information from a select number of parliamentarians or MPs and finding what they're actually talking about. So color coding and representing what these parliamentarians were talking about on Twitter as a way of tracking policy discourse and things like that. And that was live. Yeah, it is live, I think. And I think that there's also an important element of it that one of the most powerful parts of data visualization is it allows us to tell complex stories in still effective ways. And one of the ways that you can do that really is through layering information. An example of that that we had was a really powerful diagram from the Czech Republic called Mapping Czech Crime. And you could, it kind of presents a general view of the Czech Republic and what sort of crime is happening where. But you can then drill down and you click into different areas and this knee chart pops up with the different types of crime, whether they were solved or not, and things like that. So they also have that, then you could also click through into timelines and see how trends were shifting over time. But it was a really powerful way of one website telling a lot of different stories for different audiences as well. So I think that I'm going to go ahead and leave it there. But that's a quick introduction to data visualization or more appropriately, information visualization.