Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Episode #86: Mona Chalabi
177 Plays8 years ago
Welcome back to the PolicyViz Podcast! On this week’s episode, I’m very pleased to be joined by Mona Chalabi from the Guardian. Mona, if you don’t already know, is involved in a number of exciting projects at the Guardian and...
The post Episode #86: Mona Chalabi appeared first on PolicyViz.
Recommended
Transcript
Introduction and Guest Overview
00:00:11
Speaker
Welcome back to the Policy Viz Podcast. I'm your host, John Schwabisch. Thanks again for tuning in. I hope everyone's having a lovely spring and not in pollen hell as we are here in DC. I am very excited for this week's guest. She has been all over the world doing a number of great different projects, trying to visualize data, number statistics in a lot of very different ways and engaging ways. I'm excited to have with me today Mona Chalabi from The Guardian.
00:00:41
Speaker
Mona, welcome to the show. Hi, John. Thanks for having me. How are you? I'm good. Thanks. Yeah, you're busy. I am busy, but you know, you're real busy. Yeah, I've been
00:00:56
Speaker
I think the Guardian has been quite generous in making me do a lot of work outside of the Guardian, which has been fun because it's meant experimenting with a lot of different kind of formats for stuff, which is really fun. So that's what I'm thinking about most at the moment.
Mona's Work and Creative Approach
00:01:07
Speaker
Yeah. So can you walk me through some of the non-standard sorts of things that you've done? So at 538 and then at Guardian and also Sesame Street, because that's kind of awesome, by the way. In case I haven't said that before, that's awesome.
00:01:25
Speaker
Thank you. I would say that my whole career in data journalism has always been focused on the accessibility of numbers and trying to make them interesting and engaging to as many people as possible.
00:01:39
Speaker
Not long after I joined FiveThirtyEight, I decided to start a regular column called Dear Mona, where readers could write to me about anything at all, and I would try to kind of contextualize their experience with numbers. Actually, originally it was called Am I Normal? And then some people in FiveThirtyEight kind of got cold feet and thought maybe that would sound a little bit judgmental. So they switched it to just kind of calling it Dear Mona.
00:02:02
Speaker
And yeah, it was really really fun actually, it was great. It brought in a very different readership to FiveThirtyEight's normal readership and we got really really diverse questions that really made me think in different ways about how to find data and how to communicate it.
00:02:19
Speaker
So things like, is it normal to sleep in a separate bed to my wife? How many people regret getting tattoos? How many times a week is it normal to masturbate? Man of urine, right? I remember that one. Yeah, volume of urine. It's funny. That one was actually one that I didn't answer and I gave it, I was doing a talk and I was listening for like comedy purposes basically. Every single weird question I've ever got like, are you wearing socks right now? Clearly just some kind of weird pervert.
00:02:46
Speaker
someone who wrote like you hear that like asphyxiation during masturbation is dangerous but it's not really is it like give me the odds that i'll be okay and i was just like not going to write you that piece sorry and then i gave the piece i gave the example of someone who writes me saying
00:03:02
Speaker
how much P is a lot of P. And then after I did the talk, people were writing to me being like, why didn't you look into that question? That's such an important question. So I ended up going back to it. And again, it's a really interesting example, I guess, of ways to
00:03:17
Speaker
make numbers that are perhaps a little bit difficult to get your head around a bit more familiar. So with things like volume, I think, you know, we're so used to just drawing a y-axis with milliliters and kind of walking away from it. But even myself, you know, I have a sense of what one liter is, but the way that I visualize one liter is by thinking of a one liter bottle of water. And so to visualize urination volumes, I use different sized drinks containers, which I guess is pretty gross when I say it like that, but sure.
00:03:44
Speaker
So like a can of liquid versus like a two litre bottle versus like a small cup. And yeah, seems to get the message across I guess.
00:03:53
Speaker
Yeah, I guess. No, it does. I mean, these non-standard ways of communicating, I think the one that comes to mind more recently from Guardian is the Vagina Dispatches piece, which won a gold medal award, well, a special award at the Milofie Conference a few weeks ago. So first off, congrats. But can you talk about that piece and especially the part where you enabled or allowed people to come in and draw?
Exploring 'Vagina Dispatches' and Its Impact
00:04:19
Speaker
And how do you think that engages people in a kind of different way?
00:04:22
Speaker
So the series itself I made with a brilliant filmmaker who used to work at the Guardian and has now left. Her name's Mae Ryan and it was a four-part video series all about vaginas trying to basically educate people of every single age group and it was incredible. We got
00:04:39
Speaker
E-mails from people with every single age group like truly like, you know 67 year old woman range by saying I just found my peehole after watching episode one and then like a 14 year old girl saying that she was gonna get labiaplasty and change the mind after watching the episodes a really really pleasing but but we wanted to find ways to
00:04:58
Speaker
engage people beyond just simply kind of sitting there and watching the series. So I can't take credit for the draw of Ulva idea that came from a wonderful colleague called Eliza who decided to design this draw of Ulva. And again, just the number of submissions was incredible. I think you might remember. I think it was like 22,000 I think was... Oh, I thought it was more than that.
00:05:21
Speaker
So we closed it, you're right, we closed it after 21,000 but we kept on receiving. And as I'm sure you know, you know, a lot of people decided to, with a wonderful sense of humour, draw penises instead of drawing vulvas or they drew vulvas. And honestly, that was kind of the point, like, part of the reason for that sense of humour is a, a bit of discomfort with the topic and b, because that's actually honestly easier than drawing a vulva, like,
00:05:47
Speaker
Part of the rationale for starting this entire series, I remember it was a conversation that May and I had really early on. When we were trying to figure out, you know, does this series really need to be made or do people actually get vaginas? And I was like, no, when I was in school, I would doodle dick pics on my exercise books. I would not have even known how to doodle a vulva. And it was a really important proof of concept because it was to challenge our readers to say, oh, okay, you think you know this thing, then just kind of draw an illustration of it.
00:06:13
Speaker
And actually, to come back to the idea of data visualization, I think that's a really, really useful tool, too. I'm sure you've seen the New York Times have kind of tried out a couple of these draw a chart things that are really, really interesting because it's so easy, I think, when you look at a chart, we're just such arrogant beings to be like, oh, yeah, I knew that. I thought I knew that.
00:06:32
Speaker
But when you're actually challenged to plot it out yourself, you're like, oh, it's really, really important to be confronted by your lack of knowledge before you're shown the information. So I think it's a really, really valuable tool. I think sometimes I guess we have to be a bit careful about when we use it because you can very much imagine
00:06:51
Speaker
People getting tired, basically, getting fatigued by constantly being challenged by these things. Well, the other thing that they do that I think is interesting for some of those New York Times pieces is they don't let you get into the story until you draw the line. And I do wonder how many people just say, forget it, I'll just move on. And I think the latest one they did, I remember there being a button that said, it said something like, forget it, just show me the answers. Yeah, I remember that, yeah. So I'm guessing there was a bunch of people who sort of just left the page and now they're allowing people to sort of, you get both pieces.
Hand-Drawn Data Visualizations
00:07:20
Speaker
The other thing you're doing at the Guardian right now is just the facts, is the hand-drawn data visualization. Can you talk a little bit about your process behind that? And also, is that another engagement strategy? And do you think sort of the hand-drawn things is the new wave of visualizations? I mean, obviously, dear data from last year was a huge success. Is that a new wave or is it going to, is it ebbing and flowing?
00:07:45
Speaker
I don't think it's a new thing at all. I think anyone who's been doing data visualizations very often, they'll start from a kind of hand-drawn thing before they move to using other tools. I'd actually separate out the data illustrations that I do from just the facts. So the data illustrations I've actually been doing for over two years now, but I wasn't always publishing them. So I have quite a few actually on my Instagram account. And part of that came to be totally fine from a real frustration of working at FiveThirtyEight. I found myself surrounded by people who
00:08:14
Speaker
were very well versed in digital tools, very well versed in polling, and perhaps sometimes struggled to communicate their ideas to a wider audience. And I was also a little bit skeptical of them, to be honest, like I honestly thought there was a little bit of arrogance at 538 about
00:08:32
Speaker
the degree of precision with which they were kind of doing their research, which I'd be happy to come back to later on. I think it explains a lot about what went wrong with the election forecasting. But anyway, so the point of the illustrations was actually to serve several purposes. One, it was to show that a human is responsible for doing this stuff, which often people can forget. The charts that appear online sometimes look like they were just generated by some kind of
00:08:57
Speaker
omniscient computer rather than, you know, a person being behind them. So I think when you see something that's been hand drawn, you can, it's impossible to forget that someone, someone did that.
00:09:07
Speaker
Because these are hand-drawn lines, it's also impossible to forget that they're imprecise, right? And the data itself that I'm very often using, you know, if I'm drawing polling data, it has a margin of error of, let's say, three percentage points. That shaky line, honestly, you're looking at it and you're thinking, oh, that could be anywhere between this and this. You're not walking away with a number with a decimal place. You're getting the overall story, which is actually what the data set is truly communicating. It's not communicating things to that degree of precision. So it was also an effort to get back to that.
00:09:35
Speaker
And it was also just kind of an experiment I guess in creativity and trying to engage new audiences. And the thing that's really nice about it is very often if you publish a piece there will be comments beneath the piece and perhaps some of them refer to the data visualization and some of them refer to the writing. By using Instagram as a medium it's really fun because
00:09:54
Speaker
I see all of these questions just specifically about the data visualisation. So I see people saying, hey, you know, is this about one specific age group? Is this about everyone? Is this just about men? I don't actually understand what you're visualising here. Or, you know, in one case, for example, I remember after the death of Muhammad Ali,
00:10:12
Speaker
produce this data visualization about his sporting record. And someone just commented being like, you've completely misunderstood this data, the site that you took it from, actually uses not his sporting record, but the record of each of his opponents prior to fighting him in order to gauge, you know, I just completely misunderstood the data. And it was fantastic. I got to like, engage with that person, take down the chart, you know, like, it's a really, really fantastic tool for
00:10:38
Speaker
just figuring stuff out and not just kind of like having a conversation in an office and putting it out and then kind of walking away from it. Right. So one of the things you said was really interesting was this idea that it doesn't appear as precise. So I think that leads to a whole other question or maybe Pandora's box about visualizing uncertainty.
00:10:56
Speaker
Right? And whether the field of data visualization has addressed that appropriately and whether we think people actually understand uncertainty, understand percentiles. And I'll hold on to my opinion first, but I'll let you go. I mean, so I guess my question to you is, is there a way to educate readers about uncertainty, about percentiles, about distributions that doesn't feel condescending or overwhelming? I mean, I think overwhelming is probably one of the key things.
00:11:26
Speaker
So the first thing I would say is that I would argue
00:11:31
Speaker
pretty much everyone understands uncertainty. I think it's actually a large part of the reason for a growing distrust and dissatisfaction with national statistics, right? There is a large swathe of America that doesn't trust the BLS's unemployment figure, and it's partly because they understand uncertainty. They understand that no data collection technique can possibly measure national unemployment to a decimal place. However,
00:11:57
Speaker
Other people have taken that, for example,
Communicating Uncertainty in Data
00:12:00
Speaker
then candidate Donald Trump saying the figure could be as high as 39% or 42% and gone way, way too far with it. And again, to take some responsibility for that, that's partly because people like me have spent years just communicating that top line figure without saying, you know, the unemployment rate is somewhere between 2% and 4%. Why do we think that readers are so lazy to not want to hear that? Between 2% and 4%. I think that actually is the basic building block
00:12:27
Speaker
of communicating uncertainties to begin with a range, right? We know that most data sets actually do exist around range. Let's just communicate that range. And, you know, again, it's interesting to me how much I feel like we've almost regressed on this point. Like I remember when I was younger, seeing so many charts that were kind of two sets of dotted lines and knowing that the value fell somewhere between that. And I see that so rarely now.
00:12:52
Speaker
Yeah, for one reason it's convenient, right? It's 3.2, that's convenient. So let me give you one example that's not necessarily a lay audience, but one example of where uncertainty could potentially cause a problem. When I worked at the Congressional Budget Office, we had an outsider come in once, an academic economist and say, you all should really be
00:13:11
Speaker
presenting the uncertainty around each of the scores for any, you know, for any bill you produce a score for, you should be presenting uncertainty bounds. Right now, 10 years out on any bill, the uncertainty is essentially plus or minus infinity. I mean, we all sort of know that, right? But you set up some model, you do the best you can, and you come up with an estimate.
00:13:32
Speaker
The pushback on that argument of let's give members of Congress an uncertainty bound around this dollar number is the argument then goes around the uncertainty, around the confidence interval, and is no longer about the policy. So we know that the $600 billion is wrong. It's wrong one way or another. Let's argue about the policy and whether it's good or not. And if we're in the ballpark of $600 billion, good. And if we're not in the ballpark of $600 billion, we did something wrong. We learned from it.
00:14:01
Speaker
In that particular case, the uncertainty actually does a disservice to the audience because the debate then becomes not about the most important thing, which is the merits of the policy. Does that make sense? It does, but are you certain that that's the way the debate would go?
00:14:15
Speaker
I actually, I could imagine a completely different type of debate where, in fact, actually what ends up happening is, and this is another argument against it, so I'm kind of arguing your point, I guess, is that everyone who is in favor of that particular policy, say, says, it will cost as little as blah blah, and everyone who's against it will say, it will cost as much as, yeah. So you have this spread, and now you can sort of cherry pick where you want to be in this confidence interval, whereas if I say to you it's $600 billion,
00:14:41
Speaker
you just have to accept that number. But you don't have to, that's the thing though, that's what I think people find so frustrating is this ultimatum of, hey, this is the number, either accept it or don't. And I honestly think as well that again, people are smart when you or I hear it could be as little as, you don't forget those words, or when you hear it could be as much as, and I don't think that's because we're some kind of like,
00:15:00
Speaker
Data gurus I think that's like really basic common sense actually everyone listens out for no I think I think that's right and like I said I think that's a pretty Specific example and the range may be the way to go But I think people like to be able to have a number because then again if you have the unemployment rate a year ago Today and those confidence bounds overlap do people understand what that means and when those ranges overlap I mean, maybe they do maybe they do understand and maybe they would understand it better than if we said it's this percentile You know the percentile went from this to this
00:15:31
Speaker
Yeah. I mean, obviously this is a case by case thing. It can't be dogmatic. But let me give another example, which is to come back to FiveThirtyEight, right? Everyone, I'm sure, pretty much everyone that's listening to this podcast is very familiar, can visualize that homepage because they relentlessly checked it in the weeks leading up.
00:15:48
Speaker
to the big day. Before you even get into the question of range, why on earth were there decimal places on the probabilities at the top of the page? Completely, completely misleading. And to say, oh, you know, readers don't understand probabilities. That's why we're being criticized. No, you as journalists have a responsibility to communicate the information in a way in which it is least likely to be misunderstood. So that was the first error.
00:16:09
Speaker
Second of all was, were they even visualizing the right information? Wouldn't it have been more powerful to have said, we think that Clinton will win between X and Y? Oh my God, I want to say congressional votes. That's not right. Electoral votes. Electoral votes. Oh God. It was so obsessed. It was so, so long in it and just had a complete... We move on to France. We're moving on to France, so you know. That's exactly why I heard that. The whole system. And the UK after that. Yeah, yeah, yeah. So yeah, and I understand that, you know,
00:16:38
Speaker
actually electoral votes are really really confusing for but you know what that's the reality that's the way politics works in this country and it doesn't work by saying this candidate has an x percent chance of winning and even if you were to do that like let's say
00:16:53
Speaker
I don't know if you could have really communicated the range, but to say how it could have gone either way, if you think of probability distributions, it's really interesting. I remember at OpenViz conference last year, someone was talking about one way to visualize uncertainty and they were talking about these probability distributions and that one of the oldest ways that probability distributions were described to children was just dropping balls and showing how they stack in different places.
00:17:17
Speaker
I guess you could have said like here are 10 scenarios for the world on November 9th and each one of these balls represents those scenarios. I don't know like I still don't think there was enough creative thinking partly because honestly again people like you and I are incentivized to overstay accuracy. It's how we get our jobs, it's how we get our credibility, it's how we get our like kudos being like look how precise we can be and honestly like a lot of it is just deceitful
00:17:45
Speaker
No, I agree, and I think when someone says candidate X has a 10% chance of winning, even most people's first inclination, first intuition is to say, oh, okay, or that team's not gonna win. Exactly, exactly. And that's probabilistic, that's not true. But I think there is a little bit of hindsight in this when places like FiveThirtyEight and The Times and whoever, whoever's doing the forecast, they come back and say, well, I said there was an 8% chance
00:18:11
Speaker
that he was gonna win, or she was gonna win, or this team was gonna win. And so I said it, and so therefore you should have understood it. But I don't think that's how most people, data people are not, intuitively think about when they see these numbers.
00:18:27
Speaker
But they know that, right? Like Nate knew that's how people think it's right. He kind of, I guess, took a bit of a gamble in that the fact that people misunderstood it wasn't a problem in previous years and hopefully it wasn't going to be, you know, I mean, again, you think of the weather, like if someone says to you, there's a one in five chance it's going to rain, you're not necessarily worrying about your umbrella. I don't know. It's like it's all of these human basic human behaviors that don't take that much thought to really anticipate, I don't think.
00:18:55
Speaker
Maybe that's unfair of me.
00:18:58
Speaker
No, I mean, well, I mean, obviously there's a lot of stuff going on. There's issues with the polling, which is a whole other, and maybe a conversation we should have, right? Like one thing that's really interesting about the election is that there are all these different visualizations and dashboards and what have you. And at the heart of it, they're all sort of rooted on data that may not have been very good quality. And so how do we then, just from a data visualization side, how do we assess the visualizations? It's a visualization that's super creative,
00:19:26
Speaker
But based on polling data that's not very good, is it still a good visualization? No, I think absolutely not. I would even go one step further before you even question accuracy, which is what was the utility of this? Imagine if all of those resources that had been poured into predicting who was going to win this election had been poured into informing the public about the effects of their policies, had been
00:19:50
Speaker
And of course, it's very, very difficult, very difficult to accurately assess and gauge Donald Trump's business ties. But it could have been done more effectively if that time hadn't been spent just trying to predict this. And again, you know, I did some research into this after the fact.
00:20:08
Speaker
And honestly, it's not totally clear. People don't really know. But the way that communicating polling and communicating forecasting in this way can actually influence electoral outcomes is something that has been written about extensively. The bandwagon effect or the underdog theory was the fact that people who supported President Trump thought that he was going to lose. Did that actually encourage them to get out and make that protest vote even louder? You know, even if that's just a theory to me,
00:20:34
Speaker
The mere fact that it's a credible theory is kind of worrying. It really is. Yeah, no, I agree. I mean, if you think that she had a 95% chance of winning, maybe you don't go vote. Right. I mean, I know for a fact, I know personally people who did not. I also know people who I spoke to over the course of the election who wanted to support Sanders, say, and were like, that's a wasted vote. I'm not going to do it. And that's also not the way that democracy works. Right. Right.
00:21:01
Speaker
And the only evidence for saying that he wasn't going to win was that polling, was that forecasting. So aside from fun balls dropping and animated things and drawing vaginas and penises, how do you think of trying to convey distributions or trying to get people to understand what the median is?
Using Analogies to Explain Statistics
00:21:20
Speaker
I'm going to just start with the median. Or the mean, right? How do you people understand that?
00:21:25
Speaker
If you have a bunch of people and you throw some basketball centers in the middle of those people, the mean of the height is gonna not be representative. Like, how do you think about doing that?
00:21:35
Speaker
Well, it's partly what you just did, which is like analogies, real world analogies. You know the thing when you're teaching kind of high school kids math, it's like they just want some application to the real world, otherwise they have no incentive to kind of pay attention. So I think real world analogies are really important. I think we need to get smarter about ways to annotate charts that aren't overwhelming, but can kind of communicate this.
00:21:58
Speaker
So you can imagine toggling between two different views, which is like the optimistic view and the pessimistic view and not even necessarily toggling. Maybe it's like a gif. You talked about ways to communicate the inaccuracy of polling. Maybe all talks about polling are visualized in pencil and charts based on the Census Bureau are done in pen. It sounds really, really silly, but people understand that pencil is something that's, you know, when you're still figuring it out and pen is something that you do when you're like a little bit more committed to an idea on something. I think there's so much.
00:22:28
Speaker
that we can do beyond what we're doing right now and i get that it's tough but i also think we have so many tools at our disposal you know i keep on saying to people about so i mentioned that on instagram you know there's this really great capacity to have a conversation kind of beneath the picture
00:22:44
Speaker
But in addition to that, I remember ages ago, Gorka used to have on their site a tool where you could click anywhere on a photo and leave your comment on a specific place in the photo. I've yet to see charts where you can make comments on specific types of the chart. And imagine if you were to hover over a chart you made and notice that actually all the comments and all of the questions are clustering in one specific place.
00:23:04
Speaker
that helps you go out, it helps you find new stories, and it also helps you figure out what you've done right and what you've done wrong about that visualization. So I think, first of all, we need to get a little bit better, I guess, at figuring out, which is something that you mentioned earlier, to what extent do people understand uncertainty already? Are they looking at the chart and asking, hmm, how precise is this? Or do we need to start off by kind of sewing those questions into people's minds, and then taking it further, you know?
00:23:29
Speaker
Yeah. I mean, do you think that the professional data visualization people, I hesitate to call them data visualizers, but okay. So professional data visualizers, are we creating charts and graphs for each other and not for people that we're actually trying to communicate with? Yeah, absolutely. I think a lot of these charts are made
00:23:48
Speaker
by experts for experts. You see the chart making process in a lot of newsrooms and it's kind of shown to another peer who also is well versed in this stuff. And it's like, does this make sense to you? Sure. How often do you grab someone else from a completely different department and say, hey, do you get this? I don't think that happens often enough. And I also think that, you know, and I've seen this in so in so many of my peers.
00:24:10
Speaker
You want to get better. You want to get really, really great at what you do. You want to hone your skills. And very often you do that by producing charts that are increasingly complex. That's how you build your career, right? I've had to really swallow my pride, basically, to produce these hand-drawn things. I think part of the good thing about them, but also part of the, frankly, humiliating thing about them is that anyone can look at that and think, oh, I can do that. That's why I want to do it. But it's also, there's no pride in that. There's no, look at this incredible thing that I produced and no one else in the world can make it, you know?
00:24:39
Speaker
Yeah, but at the same token, I remember I remember as a kid going with my mom to the art gallery in Buffalo where I grew up and saying, you know, that's just, you know, Jackson Paul is just throwing paint on it like anyone could do that. But like, but he did it like someone has to do it. Right. So
00:24:56
Speaker
I mean, you're drawing visualizations, and some of them are fun, right? Some of them are boogers coming out of people's nose, and some of them are more serious, but you are doing them, and no one else is doing them. And, you know, the same thing with the Dear Data project, right? They were not the first, obviously, to hand-draw things, but the first to do that sort of project, and now that's sort of grown and exploded, and other people are trying that.
00:25:19
Speaker
And it's honestly a super flexible tool, like I like the fact that you can turn it around really, really quickly. So when a piece of news happens, you can make it part of the conversation pretty much immediately. And actually, it's relevant to something else that you're talking about, which is not only uncertainty, but making sure that this is accessible to as many people as possible. And again, this isn't something that I don't think necessarily happens with charts as often as it should, which is
00:25:40
Speaker
iterative storytelling, right? So imagine if instead of just throwing up the chart, you draw the x-axis, you draw the y-axis. And it's really interesting we're talking about this over a podcast because, you know, at one point I had a brief kind of slot on NPR called Number of the Week. And if you say one number, if you say 6,575,832, people have just about managed to grasp that number. If I were to then quickly say to you that 7,253 of those people, I've lost you. I've completely lost you. And so
00:26:09
Speaker
getting people to visualize a chart in their mind by describing it saying you know imagine a horizontal line that goes from 1900 to 2017 and now we're going to think talk about how migration has changed over that time you know that's a really really powerful tool for not only giving people the information in a piece more fashion but also explaining a lot of those
00:26:31
Speaker
different perhaps uncertainties or inaccuracies as you're doing it. Kind of like, I don't know, at the right moment. Because it's so difficult I think for us to know where exactly people's eyes land when they first look at a chart. And by using this, you actually control exactly where their eyes land. I don't know.
00:26:49
Speaker
No, that's right. You control them. I mean, it's like reading a book, right? The book itself doesn't have to be a linear story, but you still have to read it left to right. But one of the questions is, in this world where people are scrolling quickly,
00:27:05
Speaker
Is it engaging enough when you show them the x-axis to start and, you know, maybe it animates, are they going to stick with it? I think, as an example, the Bloomberg climate change one demonstrates that, yes, people do interact with that, right? People do engage with that, but is that going to work time and time again? And I agree with you that the annotation is super important, but at what point does it come overwhelming for people to have to see the median explained on every chart? You know, I don't know.
00:27:32
Speaker
But again, it's just about developing new, I mean, we're assuming obviously all of this stuff is on a screen. I think that's a pretty safe assumption. And a screen has a whole load of possibilities. You could just have like, you know, just one toolbar consistently throughout that is your basic statistical concept. Let's say you hover over.
00:27:51
Speaker
the line, which is the median, a thing flashes up on the side, which is the median thing, if you want to have the median explained to you. And so as you're moving around the chart, all of the statistical concepts are kind of there if you want them. There's so much that we can do, but we're just not really doing it. I'd also say that, as we're talking, I'm just sort of maybe revising my philosophy a little bit. I've been sort of saying, people don't understand the median. They don't understand this. That's not totally true in the sense that
00:28:18
Speaker
a person may, over the course of reading graphs, now understand the median. The New York Times, for example, has demonstrated that people understand what a scatterplot is. They have started doing scatterplots pretty regularly. So that's now becoming as familiar as line charts and bar charts. So if people can expand their graphic literacy, then maybe their statistical or their numeracy can expand as well. Maybe it's just a longer process of getting people to understand more complicated statistical terms than graphs. I don't know.
00:28:48
Speaker
Yeah. Okay, but one more big thing I wanted to ask.
The Power of Data Visualization in a Post-Truth World
00:28:52
Speaker
In this world of alternative facts, in this world of a post-fact world, are you an optimist or a pessimist when it comes to using data, using visualizations to inform people and change people's minds? For the good, let's just say.
00:29:09
Speaker
I'm an optimist as long as I think it's really important that I remain open-minded about the possibility of having my own mind changed.
00:29:19
Speaker
So, you know, this just the facts column, which you mentioned earlier on, there are so many fact checking columns around. And when the Guardian asked me to start writing one, I was kind of reluctant because I feel like they're not really, I don't know, there's just so many out there. I don't know how much one more will add any value. And part of the reason why I'm optimistic is because I tried to write this in a way in which
00:29:40
Speaker
someone feels like they are on a level with the writer rather than being preached to and I try to talk through every single step of my process like I'm really trying to not be patronizing literally just google this use file type colon xls if you just want to find data results and look at the third page on google don't just stop at the first or second page you know all of these sort of things
00:29:59
Speaker
doing a kind of reverse search where i look for a chart by doing an image search and then i'll follow that chart to find the original data and i really really hope that that is maybe starting to change people's minds i don't know i also think to come back to something i guess we've kept on mentioning is that i don't think there are any hard and fast rules about any of this i think it's all about
00:30:20
Speaker
Constantly trying to figure out, do I need annotations on this chart? Is a visualization even the best way to do this? Maybe this should be an audio chart, which is something I've been trying to experiment with lately. Maybe this should be a hand-drawn chart. Maybe this should be an interactive. And I think that as long as we continue to experiment with those things, we'll continue to hopefully be reaching new audiences.
00:30:41
Speaker
I don't know. I've mentioned my inbox a couple of times, and I think that's actually one of the least used tools by people who work in data visualization. I think we need to be asking people who don't necessarily agree with us what questions they want us to ask and why they don't necessarily agree, and hearing those thoughts with a really, really open mind. Yeah. I think that's right. I think that's super important.
New Ventures: 'Business of Life' on Vice
00:31:02
Speaker
I want to give you a moment just to talk about you have a new venture.
00:31:06
Speaker
I mean, you don't have enough going on. Just keep adding. Also, I want to hear later on, I'm excited to hear about the sound charts because that's going to be fun. But new TV show on Vice. Yeah, it's called Business of Life. And it's a panel show, which is kind of structured around statistics. So we end the episode on aging. Last week, it was actually Sunday night.
00:31:29
Speaker
And for example, we'll talk about how the median cost of a funeral has risen in America, and then the experts will kind of weigh in. But it's nice because the conversation is kind of structured around these statistics. Even if people might hugely disagree with each other, there's kind of this one starting point.
00:31:49
Speaker
Yeah, so they all agree as they come in that this is going to be the ground rules of the data as it were definitely question the numbers but in a way that was really fruitful as well again for like I guess getting people to think of it more critically so like an expert would say okay this is like the median but actually we know that in certain parts of the country if you know is a way cheaper or
00:32:09
Speaker
You know, this statistic that you've put up here includes like the most basic package, but actually, if you go for a cremation, you can save a load of money. I don't know, it's just a really interesting way to kind of start the discussion. I guess that's kind of a weird example to give from the show. We talk about all kinds of things, but the funeral cost statistic particularly sticks in my mind. And how many episodes can we look forward to?
00:32:29
Speaker
There's 14 in total. 14. All right, so we've got a full lit. Milo's picked up. All right, great. Yeah, yeah, yeah. All right, Mona, thanks so much for coming
Conclusion and Farewell
00:32:37
Speaker
on the show. This has been a lot of fun, and I look forward to watching the Vice series and seeing all the other stuff that comes out soon. Thanks so much for having me, John. Take care. You too. Thanks, everyone, for tuning into this week's episode. I hope you've enjoyed this week's episode and the last few, and of course, more coming up through the end of June. So until next week, this has been the PolicyBiz Podcast. Thanks so much for listening.