Moritz and Enrico on Books, Data Literacy, Their Projects, Etc.

Hey, Enrico, what's up?

Hey. We have a new sponsor here.

It's true.

Starting from this episode, mixing things up. Yeah, it's good.

Yeah, it's good. A little diversity is always good.

It's a good feeling.

Yeah. So this time, data stories is brought to you by click, who allows you to explore the hidden relationships within data that lead to insights that ignite good ideas. Let your instincts lead the way to create personalized visualizations and dynamic dashboards with Qlik sense, which you can download for free at Qlik.de/datastories . That's Qlik.de/datastories.

Yeah. So this time, data stories is brought to you by click, who allows you to explore the hidden relationships within data that lead to insights that ignite good ideas. Let your instincts lead the way to create personalized visualizations and dynamic dashboards with Qlik sense, which you can download for free at Qlik.de/datastories . That's Qlik.de/datastories.

Don't forget the data stories part because.

That's the best part.

So we have data stories number 47.

We're getting old, man.

We're getting old. Getting closer to the fifties.

Yeah. We need a cabrio.

Yeah. Half century.

It's not bad, man. I just thought we go that far.

I mean, honestly, it's so much stuff. I don't remember what we have done.

It's all a blur.

Not sure yet what we're gonna do for the 50th, but we should organize something. If you guys have any idea what we should do for data stories, number 50, let us know. Yeah.

What's the appropriate way to party?

Yeah.

Party like a podcast.

Yeah. Yeah. How is everything?

Yeah. Yeah. How is everything?

Good. Good. Back. Yeah. The year is, like. Has kicked off, and I'm starting to juggle my responsibilities again.

Same here.

Yeah. I have a new plan. So I always come up with these great plans, how everything can be much easier and much better. My new plan is I do three thirds. Like, one third is teaching workshops, data stories, these types of things. Conferences. Yeah. It's a third of my time. One third is, like, solid projects, you know, predictable stuff.

Yeah.

Craftsmanship. And one third is crazy shit.

That's interesting, because now. Now that you mentioned thirds, it's pretty much the same thing for me because. Yeah. Because I do teaching, I do research, and I do grant writing.

All right.

Yeah. These are my main three responses.

And what is your crazy stuff?

Everything for nothing.

The madness is spread all out. Yeah.

It's spread everywhere. Yeah.

Yeah. And I sort of, sort of defined that over Christmas that I actually want to aim at having, like, an equal balance between these parts.

Yeah. Good luck. Good luck with the crazy part.

Yeah, it's. Most of the time it's not sweet.

Yeah. Yeah. You know, the other day, I met one of my colleagues on the bathroom here, and he said, you know, during the holidays, I've been coding, so.

But I also. I also enjoyed coding, so I did, like, one or two projects now on my own again. And it was nice, too, like.

Yeah, of course.

Yeah. Just to write code.

Yeah.

But then when it comes to debugging and making it actually work. I get it.

Yeah, yeah, yeah, yeah. So I think we actually forgot mention that the special guest for today is us. I don't know when is the last time that we had a one to.

Yeah, yeah, yeah, yeah. So I think we actually forgot mention that the special guest for today is us. I don't know when is the last time that we had a one to.

One, but it's like 40 episodes ago or something.

Yeah, or something. Yeah, yeah. I think we had one where we had a review of the past episodes or episodes of the last year, something like that.

Okay.

And then, yeah, before that, I think the previous one was just at the very beginning, me and you talking about, I don't know, shit.

The good old times.

The good old times. Yeah, yeah, yeah. But I think, so we want to do this kind of thing more often, right. And try to talk about things that normally we cannot talk about because we have guests and we cannot really totally control what we're going to talk about. But there are a few things that we want to talk about that we are interested in. So I think this is an opportunity to basically put these ideas into the. Into the podcast. And I think one thing that we should do in the future is also ask for, do some sort of Q and a kind of thing. So if you guys have any, anything, any questions or anything you want us to talk about during this special kind of shows, we will be very happy to talk about it.

Yeah. Naturally, we're also now planning to hear a bit and thinking about who to invite and, you know, what to cover, how to move the podcast ahead. And so, yeah, your input is appreciated.

Yeah.

Just saying.

So where do we start? Let's talk about books. Right. We want to talk about a couple of books.

Yeah, I mean, we did that last time a bit already.

Yeah.

But I just received a new one which I thought it could be nice to talk about. It's called new challenges for data design. And the contributor list is really quite nice. So it's run by a, the editor is a french professor. He's called David Bjarnic, and he put together, I think, a quite nice list of people. So half of them have been on data stories already, I would say.

So.

There's Santiago Ortiz, there's Kim Rees, there's Georgia Lupi. There's Wes Grubbs. Oh, we didn't have Wes grubs. Alberto Cairo, Stefan Thiel, Jan Willem Tulp, Stephanie Posavec. Lots of good people.

Nice.

Yeah. And each of them has a chapter. Many of them are interviews, others are like case studies or overview chapters. It's a really nice, I think, status quo description of data design and so on. The downside is it's super expensive. And I sort of saw something like that coming when I was talking to the editor and asked them, like, okay, it's a Springer publication and usually, you know, these academic publications are super expensive.

Don't tell me.

And he said, yeah, it's gonna be like 60, €70, you know, but. But now it actually turned out over 100. Wow. Yeah, it's really expensive.

How come?

I don't know. I don't know. But I will put a link into the show notes because there's a cheaper ebook version and you can also have the ebook printed again, you know, probably the print quality is not that great or. I don't know about that, but that's a way at least to get to the contents for a bit cheaper. But the book, it looks really great. So it's a bit sad that these things have to be so expensive.

So are you allowed to publish your chapter independently on the web?

I'm not sure, but I did it.

Yeah. So we should put a link on your chapter.

So there's an interview with me like on my design process and what I find important and things like that. And yeah, I just published that because I think information should be free.

Yeah. But I think that you should check because I think it's pretty much the same with research papers. Of course there is a copyright on top of them, but the author can independently publish a preprint on the web. So it might actually be similar.

Usually it's. Okay.

Yeah, yeah. And.

Yeah.

So what is, how do they define data design in the book?

I'm just curious. I think it is.

And what are the challenges, by the way?

Maybe I should have read the book first.

Yeah, yeah.

But it is sort of about these artistic design y ways of approaching the challenges in opportunities of big data. I guess that's like the general idea behind the book. So it's a very, very design heavy book, and especially given that the Springer publication. So that's kind of nice. And I think it's mainly the editors. David, pianist's initiative. I don't think Springer came up with the idea, but I think it was more. He said, we need to write this book. And I really applaud that.

Yeah.

And that ties nicely to our general theme. We discovered that academics and practitioners might get closer after all.

Yeah. This is what we briefly discussed during our last episode, but I would love to talk about it a little longer because I think there might actually be a little bit of a trend, and that's great for me. I mean, most of the stuff that I've been doing on the web, like opening a blog a few years back and even starting data stories, I think partially is about trying to connect, also trying to connect these two communities, because I think that, I mean, researchers and practitioners help each other, is crucial to make visualization thrive in general. Right. And I think at the very beginning, what really surprised me is that not only practitioners don't necessarily know what is happening in academia, but even more so the opposite. Academics didn't really know exactly that there were lots of really interesting practitioners doing fantastic stuff. Right. And I think one interesting thing is that this is changing. And it clear when we went to this conference together that now academics love stuff that is done by practitioners. And at the same time I'm sure that we are producing new research that can be applied in practice by many designers. So I think that's really good. That's really, really good. I don't know if this is. I think there are a few signs that show that this is happening. I don't know if this is going to last, but I think something is happening right now that's really, really good.

Yeah, I think you're right in that it just took a while until the mutual respect and the mutual awareness was there, that sort of now this dialogue can happen. I think maybe a few years ago, many designers were a bit more repelled, maybe by the whole paper business and the complicated words and so on. And also the scientists were maybe a bit looking down on the. Oh yeah, he's doing colorful stuff. But I think given that, or seeing now all these two fields over the years, I think both sides have realized that there's a lot to be gained. Right.

I fully agree. Yeah. And I think this is also partially due to the fact that at least from the academic side, what is changing really, really fast is that now many researchers have a presence on the web and they are really trying to curate as much as possible their work. And it didn't used to be like that even a few years back. I mean, it used to be my main task is to publish a paper in the top journal or conference. Right. And today this is no longer enough. Right. I don't know. Probably it's not happening only in this. Probably it's happening everywhere, but I can clearly see it happening in my community. It's no longer sufficient just to publish a paper in the top conference. You want to make sure that you're doing some quote unquote, kind of marketing of your research because otherwise it's not enough. Right. And some people are doing this thing very, very well. Well, especially the youngest ones. I mean, Robert has been mentioning this during our last episode for a bit, but I want to iterate this thing. Some really young researchers have been giving fantastic presentations, preparing pages on the web, posting a video, making sure that everything is available and reachable on the web. It didn't used to be like that, even a few years ago. So I think this is great. I hope it's gonna last, but I think it is gonna last because otherwise you won't survive.

Right?

You want to show, I mean, publishing a paper is not enough. Now you have to make sure that people use your work, right. So I think that that's a very important trend.

But I think it also depends on what type of research you do, right? So some tools or concrete, like toys, you know, they lend themselves much more, or frameworks, you know, to being immediately applicable and other things, more basic researchers, of course. But I mean, everything can be, I don't know.

Yes and no, right?

Yeah, probably.

Yeah. Yes and no. Because I can tell you at visa I saw at least a couple of experiments, presentation on experimental work that can be probably applied very easily on real world problems. Very well attended. Very well presented. So I think that's, that's very interesting. It's not just publishing new tools or libraries or applications, whatever. I think the experimental part is very interesting. The experimental part is very interesting as well.

Yeah. So it might be a mindset thing.

Yeah. Let's see. I mean, let's see next year what happens. But I think it's good. It's really good.

Sounds like.

Yeah. Then I wanted to talk about. So, going back to books, a few days back, I posted a brief blog post on this book that I'm reading is called statistics as principled argument. And I want to briefly mention that because it also introduces something I wanted to talk about that is data literacy. That is something we, briefly, me and you also discussed briefly in Paris. And I think we should actually focus much more on these kind of problems. But let me talk about the book first. So this book has been suggested to me by Alberto Cairo. The first time I think we were at this when he mentioned this book. And for a while I was a little skeptical. Then I finally decided to buy it and read it. I didn't read it all yet. I think I'm halfway through. But it's amazing because it's basically. So this book is written by Robert Abelson in the mid nineties. Okay. And the book is not at all about numbers or formulae or anything like that. There's very few math inside, and it's mostly about the idea. So the main argument of the author is statistics, is about providing an argument. So it doesn't matter exactly how much math you use, how many numbers you use. Every single researcher who is actually commenting or doing statistics on top of data is basically providing his own or her own argument. Okay. So it's not just numbers. And I think that's really interesting and very unique because I never really read anything like that before. And I think it's a super important, important message both for people who do research and generate these numbers and the statistics, as well as for people who read these statistics. Right. I think this is especially important, more important now that we see, for instance, in data journalism is so much more common to provide arguments through data, numbers and stats. So I think it is very important for people to understand that if a story is complemented with statistics and numbers, it doesn't make necessarily the story truer. Okay. It depends on the way you present this stuff. And he makes this argument really, really well. The book is split in a few chapters. It talks a lot about rhetorics and how to use rhetorics in statistics. So if you are listening to the show and are interested in that, I really, really strongly suggest you this book, because it's very unique. I read many other books, statistics book, that focus mostly on methods, how to do this, how to do that, and. But this is not about methods. It's about how do you actually provide an argument. Right.

Yeah. Then I wanted to talk about. So, going back to books, a few days back, I posted a brief blog post on this book that I'm reading is called statistics as principled argument. And I want to briefly mention that because it also introduces something I wanted to talk about that is data literacy. That is something we, briefly, me and you also discussed briefly in Paris. And I think we should actually focus much more on these kind of problems. But let me talk about the book first. So this book has been suggested to me by Alberto Cairo. The first time I think we were at this when he mentioned this book. And for a while I was a little skeptical. Then I finally decided to buy it and read it. I didn't read it all yet. I think I'm halfway through. But it's amazing because it's basically. So this book is written by Robert Abelson in the mid nineties. Okay. And the book is not at all about numbers or formulae or anything like that. There's very few math inside, and it's mostly about the idea. So the main argument of the author is statistics, is about providing an argument. So it doesn't matter exactly how much math you use, how many numbers you use. Every single researcher who is actually commenting or doing statistics on top of data is basically providing his own or her own argument. Okay. So it's not just numbers. And I think that's really interesting and very unique because I never really read anything like that before. And I think it's a super important, important message both for people who do research and generate these numbers and the statistics, as well as for people who read these statistics. Right. I think this is especially important, more important now that we see, for instance, in data journalism is so much more common to provide arguments through data, numbers and stats. So I think it is very important for people to understand that if a story is complemented with statistics and numbers, it doesn't make necessarily the story truer. Okay. It depends on the way you present this stuff. And he makes this argument really, really well. The book is split in a few chapters. It talks a lot about rhetorics and how to use rhetorics in statistics. So if you are listening to the show and are interested in that, I really, really strongly suggest you this book, because it's very unique. I read many other books, statistics book, that focus mostly on methods, how to do this, how to do that, and. But this is not about methods. It's about how do you actually provide an argument. Right.

Yeah. And especially, I mean, if you're like a positivist or so in statistics, you would assume statistics is there to eliminate rhetorics, you know, because your findings, you know, are independent of rhetorical tricks, you know?

Yeah, yeah, yeah, yeah.

But I totally agree. It's, a lot of it is about like discourse and arguments and rhetorics. Yeah, absolutely.

So let me take some quotes from the book. I have it in front of me because it's really interesting. So he says meaningful research tells a story with some point to it, and statistics can sharpen the story. Or interestingness seems to have to do with changing the audience belief about important relationships, often by articulating circumstances in which obvious explanations of things break down. So that's really interesting. And there is also a chapter on detecting fishiness.

Bullshit detector.

Bullshit detector.

Important to have.

Yeah, important to have. Yeah, yeah.

That sounds really, really good, actually. Yeah, good.

It is really good. It is really, really good. And the surprising thing, this has been published in the nineties. I don't know why I've never heard of it. So it's really, really good. Right. And I think somewhere there is also a chapter, at least a paragraph about the dreaded p values and how not to count too much on that. And the crazy dichotomy of something is important only if the p value is less than something. If it's not, then it's not important, which is, of course.

So let's run the experiment again until it is.

Yeah. And people are chasing the p value all the time. Right. I've been doing it myself, so I know what it's talking about. So it's really interesting. And I think this introduces also the general problem of data literacy that we discussed briefly when we were in Paris. I think especially, I especially like the title of your tutorial. It was everything but the vis. Right. And I think this is connected to that because, I mean, there are many other things that are important other than encoding information visually. Right. There is so much more going on before you are able to do that. And I'm sure you have been experiencing this problem all the time with your projects, right?

Yeah. And people are chasing the p value all the time. Right. I've been doing it myself, so I know what it's talking about. So it's really interesting. And I think this introduces also the general problem of data literacy that we discussed briefly when we were in Paris. I think especially, I especially like the title of your tutorial. It was everything but the vis. Right. And I think this is connected to that because, I mean, there are many other things that are important other than encoding information visually. Right. There is so much more going on before you are able to do that. And I'm sure you have been experiencing this problem all the time with your projects, right?

Sure. And I mean, it's also a point I like to make in teaching is that, you know, for, you know, we discuss a lot how to use the right colors and if area is better, length and, you know, all these things. And then at some point in the course, I often say, like, but listen, it's not that important. You know, it's, you know, it's good to know all these things and not to embarrass yourself and not to mislead, like, you know, blatantly. But end of the day, you know, it's much more important. Like what's your general approach? Like, which type of how do you organize your own truth finding process? And how do you verify you're on the right track? And how do you verify that what you're communicating is actually what, what arrives at the audience? You know, what is the concepts that are being formed in the audience's mind by the things you do. And this is, in the end, what counts and if these are accurate and truthful or not, and not if you pick. Yeah. I mean, yeah. You know what I mean?

Yeah. And I think, I mean, there are degrees of truthfulness, right. It's not that something is, I mean, there are cases where something is clearly wrong. Right. But these cases are rare.

Yeah. And there's also very different, let's say granularity levels you want to take into account. Right. And some of them might be good in one situation or for one audience and another one might be good for another audience. And it's very, it's a very complex activity, actually. And again, I see much more and more the link to journalism because they have to cope with the same things. You know, they dive into these big stories about, you know, complex relationships and at the end of the day, they have to write an article that people get and that people read from beginning to end. But they make sense. Right. And. Yeah, so I see this more and more, this, this general thing. And coming back to data literacy, I think that also means if we want to improve visual literacy or data literacy, I think we need to teach people how to create data visualizations, because only once you created one, you know what the wiggle room is actually, or what the crucial decisions are or how much you can actually change things up just by choosing a different chart type or choosing a different data set or leaving out a part or merging two numbers into one. Things like this.

Yeah, yeah, yeah. This is the important part because it's not just about visual representation, it's about, it's mostly about deciding what to represent in the first place, which is not a trivial decision at all. And I think this is also connected to what I was mentioning before, the degree of interestingness. I think it's mostly, I mean, a very interesting visualization is mostly most of the time interesting because of the data, the information that is presented there, not necessarily the way it is presented. Right. So I think it is very important and related to that, I'm not sure exactly how to teach that or how to. So if a person comes to me, let's say after these episodes, sends an email to us and says, I would like to become more data literate, what should I read? Where should I go? It's not clear, right? Or do you think it is clear? Where would you start?

Yeah, yeah, yeah. This is the important part because it's not just about visual representation, it's about, it's mostly about deciding what to represent in the first place, which is not a trivial decision at all. And I think this is also connected to what I was mentioning before, the degree of interestingness. I think it's mostly, I mean, a very interesting visualization is mostly most of the time interesting because of the data, the information that is presented there, not necessarily the way it is presented. Right. So I think it is very important and related to that, I'm not sure exactly how to teach that or how to. So if a person comes to me, let's say after these episodes, sends an email to us and says, I would like to become more data literate, what should I read? Where should I go? It's not clear, right? Or do you think it is clear? Where would you start?

That's a good question. No, I know there's no single place, I guess, where you.

Single place, right. Because of course you can say, well, read some stats book, read all the vis classics, read this, read.

That's the data journalism handbook, which gives at least a broad.

Oh, yeah, that's a good. Yeah, yeah, yeah, yeah.

But, yeah, but I think that's also because you cannot just read, you know, a couple of things and then, you know, and then the topic is sorted. But as I said, I think it's something you have to experience yourself, so you have to measure something you care about yourself and then think hard about how to find a good visual forum for what you have measured and what you find to be a good shape of the data and things like this. So I don't think you can get become data literate without creating or working with data yourself. That's my strong belief there. And I think that's what we should teach.

That's true. That's true. But at the same time, I mean, one could also imagine that in general, we would like to have a society where people are more able just to read this stuff correctly.

Yeah, but I think we only get there if we teach them to do.

Yeah, yeah. So one little project that I always have in the back of my mind is going to some schools and teach this stuff to kids because I'm. No, seriously. I'm totally sure that they will be able to understand basic things about data and how to read charts, how to spot things that don't work. And this is a very basic kind of literacy. Right. So I think it would be totally reasonable to organize courses for in, I don't know, at least in the high school.

Yeah. Did you teach your kids already how to read, like, complex charts?

No. Did you?

Simple stuff like line charts, bar charts. We had maps. We had. Yeah, but no. No skeleton matrix yet. No tree map.

Yeah, yeah. I don't know. I fear that as soon as they will be old enough to understand exactly what I'm doing, they will totally hate it.

You think so?

No, I don't know. I'm for this kind of, you know, you want to do something against your parents. I don't know, whatever. Right.

Yeah. I mean, it's the same for me. I sort of try to keep that family life a bit separate from my professional life.

Yeah. But I can tell you I'm really surprised to see. So from time to time, I go to my kids classroom, and they already teach a lot of things very early on in grads from first grade. So they have all around the walls of the room. It's a big, big room. They have, for instance, charts, like to count how many birthdays are for a given month in the class. And it's basically each unit that makes the bar is one birthday, and there is a cupcake on top.

You could send kids out and, you know, count certain types of trees or birds or whatever. You know, it's. I think there's a lot you could do in this area, like just being in an investigator and being a detective. You know, that's something that kids like to do and figuring stuff out and like counting as well. So I think there's lots of opportunities. So maybe we should found a little data school. Data school for kids? Yeah, why not?

That would be really cool. That would be really, really cool. Now that I think about it, they also have some sort of quantified self kind of thing. They have one calendar where they stick different dots with different colors according to the weather outside.

Chocolate consumed.

Yeah. They do some manual data collection. So it's really interesting learning math.

I mean, why not?

It is perfect. Yeah.

And that's what I mean. Like, more personal data stuff. I think that's in many ways a good way to get started with that, because if it has personal relevance to sort of, you will take in the statistics and the truth finding and the sense making, you know, on the go, basically. It's not so remote. Yeah.

And I think this is a general principle that every time you do something that is personal that you feel a lot of connection to is so much easier and you have so much more interest in what you are doing. So I think this is a very general principle that works not only with kids, but with everyone, I guess.

Yeah, true, true.

Yeah. Nice data kit there.

We have it.

We should do that. We have it. We have a trademark at least. Yeah.

So we could talk about a few smaller things. So Iranic was a really nice project. I liked it a lot from Kyle McDonald. I think it's a bit older already, but I hadn't seen it. So Kyle is sort of an artist, hacker, technology experimenter. I know him for a few years now. He does great stuff. And he had an artist in residency, residency at Spotify, the music company, music streaming company. And, I mean, there's sort of mixed opinions about these intern. I say internship again, but it's a residency, of course. But I don't know how art and companies relate. That's often a bit tricky. But I think he made something great out of it, and he built an application that would track when people play exactly the same song at the same time across the world.

Oh, cool. Yeah, I saw it.

You just have this sort of jukebox radio type thing of. Yeah, the same track being played by totally unrelated people twice across the world. And I really liked it because it's such a nice and interesting way to think about this huge data set because you have to think. He basically had the whole of Spotify data probably at his disposal, and it's hard to do something meaningful with that, with all these possibilities. And he came up with this charming, simple idea. And I think in many ways, this is sort of the big challenge right now, giving all the possibilities with the huge databases and big data and lots of dimensions, lots of measurements and so on, is to come up with these poignant, smart, concise, intriguing ideas. You know, what to do with that. And often they can have a very simple core, like, it's just these two things at the same time. And I think we need more of that. And I think we can pull a lot of inspiration, obviously, from artists like Kyle, because they often have this unique approach to these things.

Yeah. And I have to say that I think this is a good time to say, I think, in my opinion, we need to have more artistic approaches to big data in general. I think it's really, really good, a really, really good idea to let artists play with big data sets, because not necessarily something good comes out of having a very precise goal, a business kind of approach. Of course. Course, this is very important. Right. But at the same time, I think that some artists have the. I don't know, they know how to look at these kind of things from a completely different point of view with a completely different kind of lens and generate stuff that is equally important. Right. And creating, in a way, emotions out of data. And I think the whole idea of creating emotions out of data is something that I really, really like, and I think it's really, really important.

This can kick emotions that, you know, can start from these. These seeds of thinking different about the same thing. And I'd also like to mention J. Thorp, who we had on the show a few episodes back, maybe a year ago.

Yeah.

And he wrote a great article on how there should be an artist in every library. So, because he's been to a conference about information sciences and. And library ship, and he also developed that sense of, okay, we should really, really work on bringing these communities together and maybe have something like a residency program in libraries. And there are a couple of these, and they could much more. And I agree, if it's set up right, these can be super interesting opportunities for both sides. I think what's super important is, though, is not to see it too much as, like, you know, just say, yeah, the artist, he's gonna do something, like, interesting, but ultimately effectless. Yeah. You know, just decoration or, you know, just repurpose the artist for, yeah, we need a new website, or, yeah, let's.

Make a data collection.

You know, but, you know, I think on both sides, there really needs to be openness to understand the other side's mode of working and what's important. Important to, you know, to the people. But then it can work. Really.

Yeah. I have to say that when something has a clear artistic value, it's so evident. Right. I mean, you don't need to judge it. It's just beautiful. Right. And, I don't know, intense. Yeah.

So I don't know these types of things. So I'm saying.

Yeah, yeah, yeah, absolutely. So you want to talk about a little bit about your project on Broadway? Because I think we didn't have time to talk about it.

Yeah, yeah, yeah, absolutely. So you want to talk about a little bit about your project on Broadway? Because I think we didn't have time to talk about it.

I think we just mentioned briefly. So.

Yeah, very.

There's an exhibition running right now in New York, in the New York public Library, and it's about photography and the selfie city team. So Lev Manovich, me, Daniel Goddemeyer, Dominikus Baur, who we also had on the show. So you see, it's a big family, it's recurring team. We developed a project especially for that exhibition, and the project is called on Broadway. And what we did was collect a lot of data along Broadway because Broadway actually spans the full of Manhattan. It's like 13 miles long and loads of blocks along Broadway. And you can collect a lot of data. And we collected a couple of social media data. We worked with taxi data. We have census data looking at. And so what we investigated is a bit like what could be a portrait of the city through the lens, of course, of that street, and also, then again, through the lens of these different data layers. So in the end, we produced this interactive installation that is like a big accordion. So in the beginning, you have the full street condensed into like an accordion squeezed together. And you can zoom into individual parts and investigate. Exactly. And also see how a whole neighborhood, you know, how the different neighborhoods compare to each other in the data, but also in street view images, Instagram images. So it's a mix of qualitative and quantitative research, again, a bit like selfie City, but for a street. And so it's now on show in the New York public library as a, like, interactive installation. I just heard it's gonna be shut down or like, closed for three months, unfortunately, because there's like. Yeah, there's like a renovation urgency, more or less. But now we are launching the website soon.

I was planning to have a feature.

I think they are gonna research group right now, more or less. Yeah, yeah. It's very unfortunate, but it will be available on the web and also the interactive thing. So we worked hard, especially Dominico's worked hard to make that work on the weapons so we can publish it. Yeah. So that's happening. That's cool. Data cuisine, which we talked about also in one of the first episodes. Really like the one on food is going strong. So we had a great edition last year in Barcelona connected to Sonar D and the Big Bang data exhibition, which was fantastic too. We also never talked about the exhibition is really good anyways, so this was good. And now we have another version we're just working on now that's actually a catering event. And it's for a group of international journalists, investigative journalists, and they have their sort of annual meeting and awards. And we make specific data dishes, you know, for each of the countries they work in and with meaningful data sets and so on. So that's very nice. And, and we plan to do more this year. So daily cuisine is going strong.

When is the next? It's actually not when, Friday.

So probably, you know, it's already past when the episode comes out in Berlin and we try to do something, let's say in April or may in Italy, but. Yeah, but I can't say much more about it. But that's like the current plan, so.

You could love to be there. Can you at least tell us where.

We'Ll be in Italy?

Okay. Okay, good. Yeah, sure. Yeah.

So that's, that's fun.

Well, the one in Italy is going to be the best one, of course.

Italy, yeah.

Yeah.

We are also having one planned in Holland, but I'm not sure about the food aspect. I would prefer the Italy edition, I guess.

Do you want me to spend a few words on that?

If you have the choice. Italy is pretty good, then. The OECD data portal came finally out of beta. That's been a big project for last year, so that went live. There will be many more developments there, but if you're interested in data, which I would have said, assume it's worth checking out. There's loads of data sets there. And the visualization there is really more about finding the right data sets. Not so much about building super fancy stuff, it's pretty basic stuff, but I think executed quite well. And what we try to do is be really smart about little child in the search results already so you can get a sense of what are the most interesting data sets straight away. So there's lots of like Sparkland charts and little nice little details.

If you have the choice. Italy is pretty good, then. The OECD data portal came finally out of beta. That's been a big project for last year, so that went live. There will be many more developments there, but if you're interested in data, which I would have said, assume it's worth checking out. There's loads of data sets there. And the visualization there is really more about finding the right data sets. Not so much about building super fancy stuff, it's pretty basic stuff, but I think executed quite well. And what we try to do is be really smart about little child in the search results already so you can get a sense of what are the most interesting data sets straight away. So there's lots of like Sparkland charts and little nice little details.

So I'm just curious to hear, I don't know if you can talk about it. Do you have any data or information about how people actually use the portal?

Not yet, not so much, because it was now mostly in beta, but we will monitor that. Yeah, it's gonna be interesting. Like how much they go into the different sections and how deep they go and things like this.

So that's gonna be mostly scientists or what? Who is using exactly? It depends. Yeah.

Yeah. So it's often it's analysts or journalists or researchers also. I mean, OCD is a general information source for wide audiences.

Yeah. Okay. I have to say that I think a few years back, it used to be a big trend of having many new data portals, and now it's almost gone. Right. I mean, there are a few established ones, which I think it's a good thing. Right. But we had this old period when a new, every few months there was a new big data source coming out. Yeah, yeah.

But it's going to be interesting. Like who are going to be the trustworthy data sources in the future? And, you know, is there a market for data, really? And you know, these types of things.

Yes.

Or in which forum is there really a market? And I think that's sort of interesting, like, especially for these general datasets. Yeah. There's also wiki data, which I'm sort of following, which could be interesting. It's like the Wikipedia counterpart. Like the data counterpart to Wikipedia. Things exist. This is all very exciting, I think.

So I think this is a good time to stop for a moment and talk about our good answer. So you've been trying the Qlikview, right?

Yeah. So there's actually two products. So Qlik is the company. And then there's Qlikview that's been around for a long time. I think more than ten years, actually, maybe 20 even. So that's enterprise software for business intelligence. Quite established, powerful data engine behind it. And it's a strong analytical tool in that context. And now there's a new tool from the same company. It's called Qlik sense. And that's a very fresh development. And what I really like about it is fully oriented towards the web. So even if you download the standalone application, it's actually like a chrome, you know, that just has a skin. But it's basically a webview that is being rendered. All the charts are being rendered in SVG and it's all native web technologies. And basically what it's good for, the Qlik sense software is a bit more approachable for Qlikview. You really need to be an expert and know the scripting language and write adapters to your databases and things like this. So it only makes sense if you have a big company and you have real it people working on that. And Qlik sense is much more a product for everybody to get started and build dashboards and analyze data themselves and really quickly build interactive visualizations. It's really nice. So I was quite impressed with. So I got a demo from one of the guys, and he really, within a few minutes, build a really complex dashboard dashboard that had really good interaction possibilities. So you can zoom all the charts. You can brush data in one chart and it will be brushed in the other ones. You can immediately drill down into datasets by selecting something in one chart and then see how it plays out in the other charts. It's really nicely done. And the charts, they scale quite well. It runs on tablets, phones, because it's all web based, basically. So, yeah, it's really kind of nice. So you can try it out on click.com, comma, click with a queue. So you have to get used to that. It's Qlik. And we'll put a link to the Qlik sense, which I would really recommend into the show notes. And if you want to see what you can do with click, there's a really nice, for instance, a quantified live demo. So there was one guy who built like a Feltrand style yearly report, and he built this whole dashboard for his whole year, all the data he has collected. It's very interactive. And so because it's web based, what's really nice also is you can build your own charts and connect them to the click engine. So if, you know, for one type of data, the line chart is fine, or the bar chart, whatever. But if you have maybe a D3 component, you want to use like a network visualization, you can, there's sort of a plugin architecture, so you can just put your own visualizations inside the Qlik sense dashboards. That's super nice. And the other thing I learned because I also didn't see it much in the wild. So you might wonder why I haven't never seen, you know, the software, it's mostly used inside companies, so it's not so much, you know, on the web sharing. Cool stuff. Used a lot inside organizations and companies. And it's a really powerful tool. And, you know, this might explain why you're like, I never heard of that. But if you're not in that world, you don't see what they're doing.

No, but I actually have heard of it many, many times.

The Qlikview is like one of the big business intelligence software.

I think in business intelligence is very well known.

Yeah. And so, yeah, check it out at Qlik derivative is. Give it a spin.

Okay, what's next?

Do we have more stuff?

We always have more stuff. So I wanted to briefly mention a couple of works from my side that might be interesting for our listeners. We just have a new paper out that is called how deceptive are deceptive visualization. And this is going to be published at the next ACM CHI's conference. That is the main conference on human computer interaction. Kai is big. Yeah. And I think the results are interesting because. So the main goal of this work was mainly to. So we started analyzing and reading all these books or resources about deceptive visualization, how to live with charts, how to live with statistics, and even Tufte has a lot of this kind of information in his books. And we were just wondering, is there any experimental work out there that just basically shows that this is true?

It's like, hold on, we never.

I mean, just scratching. Yeah, just scratching our own itch, right. I mean, this is this kind of research that is purely, I don't know, experimental, really not trying to do anything that has a practical application. Right. But I was really, really curious to see how that worked. So what we did basically was to create. So first we collected a certain number of deceptive visualizations and created different kind of categories of deceptive visualizations, how they deceive. And then we focused on a few ones that are very, very common. Like for instance, how is it called when the bar chart, the baseline of a bar chart or a line chart is not a zero, truncated. Yeah, truncated y axis. And so we collected a few, we isolated a few of them and then run experiments trying to see and created for each one, two versions, one that is deceptive, one that is not deceptive, and try to basically quantify how much distortion there is. When you ask a group of participants to evaluate the information that is in the chart. Right. And so surprise, surprise, the result is that they are very deceptive.

That's good news, though.

That's good news though. So now if you want to back it up with some research, now you, you have a paper you can point to. And so one thing that we try to do, and the results are a little bit disappointing, we try to see whether there are personal attributes that actually influence the outcome. Right? Like for instance, age or level of education, gender, data literacy. But unfortunately, we couldn't find anything. So our results show, I don't know, we couldn't find any strong relationship between personal attributes.

But that might mean it's just perceptual, like it happens on the perceptual level. Is that.

Yeah, that's, that's what I'm saying. So at least those that we tested don't seem to have a strong impact. So I originally expected to see an impact and we couldn't find any.

So that doesn't mean it is not there.

I don't know.

You just didn't find it.

Yeah, of course. We just didn't find it. Yeah, of course.

Cool. That sounds really good. And I would be interested. And.

You are preprinted.

Different types of deception you investigate sounds really good.

They are very deceptive. Very deceptive.

There's a short podcast, by the way, by Jon Schwabish, who we also had on the show.

Yeah.

And he does the Rad Presenters podcast. I think we mentioned it, and they have a short episode on when and if it's okay ever to cut off the y axis so we can look at and if people are interested. Yeah, I don't think. I think it's rarely justified, personally, but.

Yeah, I think there are a few cases where it is.

Yeah.

Because I can tell you, because we've been discussing for so long with my student about whether it is always wrong or not. I think there are. It's. It's subtler than it seems, I think. It's not just black and white. You can or cannot use it all the time. Time. I don't know. I think so.

Sometimes you have to, actually.

It's got a long digression. I just want to mention that I.

Think, you know, it's often.

Yeah.

Or if the zero is not meaningful. Exactly like zero degrees Celsius, who cares, you know, if it's Fahrenheit, it's something else.

Yeah, yeah.

Okay. Yeah, whatever. So it's not on the podcast, but it's by John Trabish. We'll find. Find out the link.

Yeah, yeah. And by the way, it's great that there is another podcast somewhere that is somewhat related to what we do.

Yeah, yeah. It's more about presentations, but they also have a few database themes, of course. So.

Yeah, yeah, yeah, yeah. So related to that, I mean, to the. To the paper that I just mentioned. We also had another paper at Viz this last year on. On persuasion. So persuasion is another. Is another itch that I've been trying to scratch for a while. And so this is another experimental work where we basically try to see whether showing an argument supported by data through visualization or tables, whether there is a difference in the degree of persuasion. Right. So, of course, there are lots of. The way we had to organize this experiment is much more intricate. There are lots of things that you have to control for. And I don't want to bore you with this stuff, but I just want to mention that. So the results of our, of our experiment basically show that under certain conditions, if you present information graphically, let's say it might actually lead to much stronger persuasion than you using, for instance, a table. Right.

Yeah, yeah, yeah, yeah. So related to that, I mean, to the. To the paper that I just mentioned. We also had another paper at Viz this last year on. On persuasion. So persuasion is another. Is another itch that I've been trying to scratch for a while. And so this is another experimental work where we basically try to see whether showing an argument supported by data through visualization or tables, whether there is a difference in the degree of persuasion. Right. So, of course, there are lots of. The way we had to organize this experiment is much more intricate. There are lots of things that you have to control for. And I don't want to bore you with this stuff, but I just want to mention that. So the results of our, of our experiment basically show that under certain conditions, if you present information graphically, let's say it might actually lead to much stronger persuasion than you using, for instance, a table. Right.

And with. Graphically mean in a chart or with enhanced. No, no, no. Okay.

In a chart. So we've been basically comparing. So you have very small data set that you show either through tables or simple charts, like bar charts or line charts or stuff like that. Right. And so apparently. So it seems to depend on the. How strong the initial opinion of a person is on the argument that is discussed by the article. Right. And so if people don't have a very strong initial attitude, they seem to be very much persuaded by, much more persuaded by the graphical illustration.

They don't have a strong.

Right. If they don't have a strong opinion, if they already have a negative strong opinion, then. And our data basically shows that they seem to be more persuaded by tables. Right. But always cautious to go through very strong conclusions because this is just one single study. Right. And this is a general statement, and I think it's important to make. It's very, very hard to conclude something from one single study, but I think it's important to have one first study in this direction that shows that something is happening there.

Yeah, that's interesting. And, I mean, it seems plausible. It's because seems like people maybe have their defenses up already. If they have a strong opinion on something, they're used to arguing about the details of the problem. And they have sort of. Their bullshit detectors are like way up, so they might be less receptive simply to something that looks like it might persuade you.

Yeah.

Interesting.

Yeah. And we also have another section in the paper that basically classifies different kind of reasons why people are or are not persuaded. Because part of our experiment was also collecting open ended comments on the outcome of the results. Right. And basically, people have been writing something about why they were persuaded or they were not. Right? And then we've been coding this information, trying to categorize different kinds of reasons why people are or are not persuaded. And this is very interesting, I think maybe even more interesting than the study itself, because it gives us an impression of what happens in reality. So some people, for instance, are persuaded no matter. So just because there are numbers there, they are persuaded. So we had numerous people just writing the. It's the number that tells the truth. Right. Or it's obvious numbers don't lie. Numbers tell always the truth. Numbers don't lie. Right. And of course you have also people on the opposite side of the spectrum is, I don't give a shit. I don't. I don't. I don't believe anything that is presented this. Right. All propaganda. Another interesting one that I think it's somewhat dangerous, in my opinion. There are people who actually never change their opinion, no matter how much evidence there is against their opinion. Right. And you can clearly detect these kind of people there because there are people who clearly write. I don't care. I don't care what you show me. I will never change evidence. Evidence, yeah. Oh, absolutely. Evidence is overrated. Yeah, yeah, yeah, yeah, yeah. And this reminds me of a book that Alberto has been been suggesting for a while that is called unpersuadables. I didn't read it yet, but it's on my tool to read list and sounds very much related. Right. So what is the psychological mechanism through which some people cannot be persuaded of anything? Right. So it's very interesting, but also very fascinating. Yeah, yeah, sure. I don't know.

With a data sculpture.

Yeah, yeah, yeah. I don't know. I don't know. Yeah. So, interesting. And unfortunately, the other day I was thinking about what is the. I mean, the implications of these two studies is very bad because people can be very easily persuaded and very easily decepted. Deceiving. Sorry. So it's not very positive what I'm doing. So next time I want to do something that is much more positive than that. Yeah, but I mean, because you can.

Yeah, yeah, yeah. I don't know. I don't know. Yeah. So, interesting. And unfortunately, the other day I was thinking about what is the. I mean, the implications of these two studies is very bad because people can be very easily persuaded and very easily decepted. Deceiving. Sorry. So it's not very positive what I'm doing. So next time I want to do something that is much more positive than that. Yeah, but I mean, because you can.

Also persuade people of the right stuff. Right. So it's, it. There's like a black thing going on, of course. Right. So.

Oh, yeah. And this actually gives me the opportunity to mention that this work is in collaboration with two of my colleagues, and one of them, she's a professor at the NYU School of Law and she's an expert in human rights. And it's really interesting because the way this work started is because basically she came to me and said, I want to learn more about visualization. I get a sense that visualization can be used to persuade people of important problems in human rights. And especially, I mean, the kind of problems that these people have is that they know that there is a new kind of crisis. They have to convince some stakeholders to take action. And whatever you can provide, whatever new tool you can provide to them to help communicate this information better can actually make a big difference. Right. So this is an example of a very positive kind of outcome that you can have with this stuff.

Sure. And then, yeah, I think that's a really interesting topic anyways, like, the whole use of data visualization in campaigning or for. For persuading people. We could have somebody who has a bit more experience with that on the show maybe once. It's because it's a whole different field, and it's whole different also. Like playing ground and playground rules also. Then in the exploratory data or in business intelligence or in, you know, whatever we do. So it's a whole different field, and it's kind of interesting. I was sort of reminded also the complexities, you know, so we talk so much about, like, global data and how to improve the world with charts and so on. And so also over Christmas, I read up a bit on inequality, and, boy, that's so complicated. You know, in the beginning, you think, you know how it works. You know, the world becomes much more inequal, the top 1%. You know, we all have this in mind, but the more you read about it, the more complicated it becomes. Like what to measure, actually. Sorry.

Sure. And then, yeah, I think that's a really interesting topic anyways, like, the whole use of data visualization in campaigning or for. For persuading people. We could have somebody who has a bit more experience with that on the show maybe once. It's because it's a whole different field, and it's whole different also. Like playing ground and playground rules also. Then in the exploratory data or in business intelligence or in, you know, whatever we do. So it's a whole different field, and it's kind of interesting. I was sort of reminded also the complexities, you know, so we talk so much about, like, global data and how to improve the world with charts and so on. And so also over Christmas, I read up a bit on inequality, and, boy, that's so complicated. You know, in the beginning, you think, you know how it works. You know, the world becomes much more inequal, the top 1%. You know, we all have this in mind, but the more you read about it, the more complicated it becomes. Like what to measure, actually. Sorry.

Yeah.

What? What? Yeah.

Are you okay? Yeah. But in the meantime, maybe I can say that I think this is pretty much true. With every kind of big problem, once you start digging into the details, it's much more complicated than it seems. Seems, right.

Yeah, that's always the problem. And that's. And I mean, and the fine. The delicate line is, of course, in a persuasive situation, you know, how do you not sweep all the nasty details under the rug, but also make, but still make a case in a compelling way? And maybe that ties back again to the book. Right. The principal argument, like, how can we learn to argue based on facts and data, but still be differentiated and human? That's what it all boils down to in the end.

Yeah. This vaguely reminds me of a book that I read a few months back, maybe one year ago or so. It's called poor numbers, and it's really, really interesting. Yeah. It's this guy from. I think he's from the Netherlands, I guess so. And he's. I guess he's a development economist or something like that. And he's basically traveling all around Africa to visit the statistical offices of all African countries. Right. These are the people who are in charge of collecting numbers about what's going on in every single country, Africa. Right. And of course, they are totally messed up. Totally messed up. But the crazy thing is that he explains that, I mean, the higher ups are taking very important decisions based on these numbers. Right. And this is totally crazy. Yeah, yeah, yeah.

But that's all you have at some point, like at some level of abstraction and. Yeah, but sure. This is how it goes.

Yeah. Everything is much more intricate than it seems. Right. When you look at the number. Numbers. Yeah, yeah. I think this is a point. I think Alberto was making this point as well in his presentation at this. Showing the political situation in Ukraine. Right. Something like that. And actually showing that if you start digging into the details, then you have a much more complex reality than just. It's black and white, it's blue and yellow. Right. It's so much. And I think maybe the point here is what one really need to learn or teach is more of an attitude. Right. It's not just a matter of learning this or learning that. It's learning an attitude. How to deal this kind of healthy skepticism.

Yeah. And also how to live with cognitive dissonance, because what can you do? And it's part of, like, that making sense, you know, is to constantly have this. These dissonances and trying to resolve them and search for evidence here and search for counter evidence there just. Yeah. Actively makes sense. And. Yeah, I think that what all of these things we discussed today actually boil down to, which is nice. Cool. Shall we?

Yeah. And also how to live with cognitive dissonance, because what can you do? And it's part of, like, that making sense, you know, is to constantly have this. These dissonances and trying to resolve them and search for evidence here and search for counter evidence there just. Yeah. Actively makes sense. And. Yeah, I think that what all of these things we discussed today actually boil down to, which is nice. Cool. Shall we?

Absolutely cool.

I think we have an hour or something.

Absolutely.

Yeah.

Yeah. That was cool. We should do it more often. You too.

Very nice. Cool. So looking forward to our next recording with another. This time, again, a very special guest. So this will be fun. Good. In the meantime, have a great time. Bye. Bye.

Bye. Data stories is brought to you by click, who allows you to explore the hidden relationships within data that lead to insights that ignite good ideas. Let your instincts lead the way to create personalized visualizations and dynamic dashboards with Qlik sense, which you can download for free at www. Dot Qlik dot de quote. Dot Qlik dot de stories.