View transcript
- Hi, my name is Nadja Popovich and I'm a graphics editor on the New York Times Climate Team. Specifically, I tend to work on what I would classify as explanatory visuals. And, the goal of my work, quite often, is to translate, visually translate science for a broad non-expert audience. I'm gonna spend the next 15 minutes talking about a project that I worked on last year and more broadly about personalization in data visualization and how that is able to help readers form a deeper connection to sometimes abstract data. So, the project was called, How Much Hotter Is Your Hometown Than When You Were Born? And, it was a project I did in collaboration with some other colleagues in the Times Graphics Department. The idea for this project really came from the fact that well as somebody who spends a majority of their time now thinking about and visualizing climate related data, I've come to realize that no matter how we visualize it, it tends to be something that feels rather abstract to a lot of our readers. I think that's true not just for Times readers, but for many people. I think that climate change can feel like something very far away in both time and space. So, this is one of the first graphics I made when I started this job at the Times. The maps use data from the Yale Center for Climate Change Communication, which has been polling Americans on what they think about climate change, various different questions for a long time now. And, the map on your left, it shows that a majority of people in the country, in most counties in the country, say that global warming will harm people in the United States. Yet, at the same time, the other map shows that not a majority of people in all those same counties say that global warming will harm me personally. And, I think these maps really show how even among people who say that climate change really will have big impacts and is a big deal, that those impacts don't necessarily feel immediate or personal. So, at the beginning of last year, I really set a goal to figure out how we could bring some of the effects of climate change a little bit closer to home for our readers using data visualization. And so, the question is, how do you do that? And so, to answer that question, I wanna take a step back and talk about approaches to personalizing data narratives more broadly. So, I think that online using interactive graphics, we are uniquely able to do something we weren't really able to do before in static graphics and definitely not in print newspapers. And, I think that we're able to make our readers active participants in a data story. Oftentimes, we're even able to put the reader at the center of their own unique story or at least a unique-ish story, in order to get across some broader idea. And so, one of the ways we do that is through location or geography. This is a example, an article from my colleagues at The Upshot. The colors look a little bit neon up there , but it's based on a national data set that looks at income mobility for children and families in different income brackets. But, the article doesn't start by showing you this big zoomed-out national data set. Like, the whole country, here are some patterns. Instead, it actually detects where you are based, I'm pretty sure, on your IP address and centers the story on this very zoomed-in view there. So, if you're in San Francisco right now and you go to this article page, this is what you're gonna get as the top of your story. And, by doing this, you're immediately brought a little bit closer to the data. The reader has a more intimate and more human connection to it. You're able to see, okay, that's San Francisco and you can immediately start comparing it to the rest of the counties around it, Alameda, Contra Costa, et cetera. A second way that we personalize data narratives is through demographics that allow the reader to really see themselves reflected in a data set. This is from a piece that I did at The Guardian for the 2014 mid-term elections, which seems like ages ago now. And, in it, you're able to click through these buttons to really see different demographic groups in Congress. So, rather than just make some charts that say, presented the percentage of women in the new Congress versus the percentage of women in the United States, which I actually did below this graphic. I really decided to center this piece around the reader. And, I think that you-centric framing gives the reader a new and more intimate relationship to this data set. Here's another example, from Nathan Yau, who, I believe, is here at this conference. The title says it all . This is a simulation of all the ways someone like you dies in America. And the you can be changed based on gender, race and age. So, lastly, I think another way we can personalize is through engagement. So, this is by letting the user become an active participant rather than a passive consumer of an article. This is another graphic that is actually a you-draw-it game, again from The Upshot, and they ask users to actively engage and take a guess at the trend for how family income predicts children's college chances. And then, you've got this reveal, which gives a kind of personal explanation of the trend that you're looking at. It tells you how you did compared to other Times readers, what part of the trend you maybe got right, what part you didn't, before really walking you through the broader idea. And so, each of these graphics which I just showed really briefly, they use one or some combination of these three strategies for personalizing a data story. So, either location, demographics and/or engagement. But, I think each of them has a similar intention, or at least from my point of view, because I'm not the author of all of those. The intention is two-fold. So, the first is to humanize data by bringing it a little bit closer to the reader's actual lived experience. And the second is to create a deeper, more active sense of engagement with the data set, and hopefully to parlay that into a deeper understanding of what you're trying to show the reader. So, these were my two intentions as well when I took on the project of personalizing climate data. And specifically, I wanted to focus on one of the most basic impacts of climate change which is heat. So, this is actually a graphic I made last year that shows 2017 was the second hottest year on record. We tend to make a graphic like this every year kinda putting into context for our readers the news that comes out usually every January of how hot that year was compared to the past. But, in whatever form it takes, you know, this is like not a very personal view. And, this is the view of climate change we get most often. It's very zoomed out, it's global. Which, is probably fair, this is a global problem, and this graphic certainly does a good job, at least I think it does, of conveying it's point, which is that overall the world is warming. But, again, I think it can feel rather abstract. I mean, even this idea of annual global surface temperature, that's just not how most people experience the world. But, you know, we're able to make this same data set less abstract even just by putting it on a map. So, this is very similar data but no longer averaged across the entire world. Instead, it shows temperature anomalies at each location. And, you can clearly see that global warming hasn't been equally spread across the world. And, effects of it aren't gonna be equally spread either in the future. So, this kind of data exists for future projections as well. This is a graphic from the Fifth Assessment Report of the United Nations Intergovernmental Panel on Climate Change and it shows, what is says up there is the annual mean surface air temperature change by the year 2100, and it's from a series of different models. And, these kinds of projections, this is again from the IPCC, another graphic, these kinds of projections are often made out to the year 2100, but when I was thinking about this personalization project, and talking it through it with my colleagues, we really kept coming back to this idea that, you know, like, that's just again something that feels quite distant. It's not necessarily something that is like the time span most of us think on, so, we wanted to experiment instead with framing this project really around a reader's lifetime. So, can we show a reader how the world has warmed so far in their lifetime? We know we have that data. We know we can do it. And then, can we also truncate it to show them more specifically how heat could change in the future within their lifespan? And then, the last thing I really thought a lot about was what exactly the right metric was to show, because this graphic, like many of the other ones that I've shown so far, really look at this like global average surface temperature or even just average surface temperature in a location. But, like I said, that's not really how people experience the world even in their own place. So, I thought back to this other project I had done the year previously, which I worked on with my colleague, Brad Plumer, and it used data from a group called Climate Impact Lab. And, what they did, is instead of looking at average surface temperature, they counted what they considered extremely hot days. So, these were days over 95 degrees, and how these could change over time. And, we ended up mapping it like this in the piece. But, I just really liked this metric. I kept thinking about it because I think it felt a lot more tangible. All of us really know what a 95 degree day feels like. It's one of those days that is just scorchingly hot, you know, maybe you don't wanna go outside. You wanna stay in the air conditioning. And, it's also at these extremes that we really see some of the greatest impacts of climate change, not at the average. So, I thought this was also an important metric. So, I got back in touch with the folks at Climate Impact Lab and asked them if they'd be interested in taking on a new, more detailed, analysis that we could base this project on. And, to my great happiness, they said yes. So, I want to now just walk you through this project, if I can figure out how . Okay, so this is the project. How Much Hotter Is Your Hometown Than When You Were Born? And, I'm gonna speed through this a little bit. But, basically you come to this screen where it's mostly blank. You get just this prompt to enter your hometown and birth years. So, I'm gonna put in my hometown which is Washington D.C., and I'm gonna take and a birth year that gives us a nice data set and we immediately draw in this line where we tell people, okay, when you were born the Washington area could expect about 13 days per year to reach at least 90 degrees. And, I wanna just point out two things, one is that this line we're drawing in is actually a 21 year rolling average based around the year that you were born. It's not exactly the data within the year because, again, we're talking about climate here not weather. And so, we wanted to make sure that if you happened to be born in a year that was like really hot or anomalously cold, you didn't get a completely skewed answer. And, we hint at this in the text in that it says, you know the Washington area could expect about 13 days. We don't tell people that it was exactly the count that year. And so, we give you just that hint and then as you scroll through the piece, we draw in the rest of the line and we tell you that today the Washington area can expect 30 days. So, you already see that within your lifetime there has been some warming. We go from 13 to 30 days. And, on the third scroll, we really zoom out and actually change the entire scale of what you're looking at on the 'Y' axis to show you that what might happen in the future. So, we tell you by the time you're 80, and we say 80 because 79 is actually the life expectancy in the U.S. and we're just rounding, so models show there could be 50 of these very hot days and then we get a likely range because, of course, you know I'm talking about the future. We don't wanna give like a false specificity. And, through this like three step process, we introduce people to this completely, I think, different way of looking at climate related data set than they're usually used to. And I think it has this real power of putting you at the center of the story that you're usually used to seeing in a really zoomed out view. And then I go on to just explain a little bit more about the data set here before we get to a second graphic, which I think is actually really important as a counterbalance to this sort of personalization. So, here in this globe, we actually take you back to that zoomed out view, the one that you're actually probably more used to seeing, to point back at the idea that this is something that is global and interconnected. And so, you see Washington already in the context of the rest of the United States. Here, we remind you what's going on. You see in the context of Latin America, et cetera. We actually here are no longer in the personal view, so we're talking about the full data set which goes from 1960 to the end of the century. And then, we take users on a little tour as well. So, we take them to the tropics. So, Jakarta, Indonesia and show how that's gonna be an area that sees some of largest absolute increases in these extremely hot days. Now it's not scrolling . Okay, well, the next slide we actually take them also to India where we explain that there are 22 million people in New Delhi and show how that is going to look. And then, in the third part, we take you back to Madrid to explain that even in temperate regions, you're gonna see really large relative increases in these extremely hot days, even though you might not see absolute increases the way that you see in a place like Jakarta. And, what we wanted to accomplish with this map here is, again, to after we've given people this really close, zoomed-in view that helps them understand climate change compared to their own personal experience, I think that really gives them a better context in which they're able to understand the rest of the data set. They have a point of reference through which they can then understand Jakarta and New Delhi and what's going on in Madrid a little bit closer in a way that I don't think that when you just give a zoomed-out map where someone kinda has to start exploring and thinking through it themselves, they just might not have that same connection. And, I think that that is something that is really powerful about this kind of graphic that really focuses on the reader's own experience rather than just presenting them with a big, zoomed-out data set. And, I'd be glad to take any questions about this project afterward at the Q & A. Thanks so much for having me.