Nathan Yau’s Flowing Data website was one of the first data visualization websites I discovered in my own data journey. With his new book, Visualize This, now out, I thought it would be a great opportunity to talk with Nathan about his work, his book, and how his own approach to data has evolved over the last several years.
The new edition of Visualize This enriches readers with modern techniques and examples, focusing on effectively learning data visualization by exploring different data types and designing for clear communication, even for those without a formal design background. Nathan emphasizes the necessity of audience-appropriate visualizations and the selection of suitable tools, all of which have changed and evolved since the first edition of the book was published in 2011.
We obviously talk about the latest book in this episode of the podcast, including Nathan’s process for creating graphics (a lot of R and Adobe Illustrator), his professional growth from a statistics PhD program to embracing full-time visualization work. We discuss the nuances of handling feedback, the differentiation between misinformation and subjective interpretation, and the significance of constructive criticism. We also touch on challenges for newcomers in the field, the need for clearer communication of uncertainty, and the potential of virtual and augmented reality.
Topics Discussed
- Updated Techniques and Modern Examples: Nathan’s new edition of “Visualize This” brings to the forefront the latest in data visualization, incorporating modern techniques and examples that cater to both beginners and seasoned practitioners.
- Learning by Exploring: The book emphasizes a hands-on approach to understanding data visualization. It guides readers through exploring different data types and designing visualizations that communicate clearly, irrespective of the reader’s design background.
- Personal Data Collection and Analysis: Nathan shares his insights into the importance of personal data collection for self-analysis, a practice influenced by his time at the New York Times. This self-exploratory journey into data helps individuals understand the nuances of their own information.
- The Challenge of Accessibility: Making statistical concepts accessible to diverse audiences is a central theme. Nathan discusses the necessity of creating audience-appropriate visualizations and the careful selection of tools that cater to this need.
- Tool Selection: With a preference for using R for data analysis and Adobe Illustrator for aesthetic control, Nathan navigates the technical aspects of creating compelling graphics. He also shares his journey with JavaScript, CSS, and HTML for interactive projects.
- The Growth of Flowing Data: Nathan reflects on the evolution of his platform, Flowing Data, highlighting its expansion to include daily emails, tutorials, and personal projects.
- Professional Evolution and Learning: From sharing knowledge during his PhD to diving into full-time visualization work, Nathan advocates for a trial-and-error learning approach. He also touches on the nuances of handling feedback and the importance of constructive criticism.
- A Practical Guide Amidst Evolution: Ultimately, “Visualize This” is a practical guide for navigating the complexities and advancements in data visualization, offering readers not only a window into Nathan’s creative process but also a comprehensive toolkit for effective data presentation.
Resources
Guest Bio
Nathan Yau is a renowned data visualization expert and the founder of FlowingData. With a background in statistics and a Ph.D. from UCLA, Nathan has combined his passion for data with his keen eye for design to help individuals and organizations understand and communicate complex information more effectively.
Through FlowingData, which he started in 2007, Nathan has been at the forefront of the data visualization field, providing insights, tutorials, and inspiration to a global audience. His work spans a wide range of topics, from personal data collection and interactive graphics to the practical application of visualization tools like R and D3. Nathan’s hands-on approach and clear, accessible writing style have made his site a go-to resource for data enthusiasts, researchers, and professionals alike.
Nathan is author of the Visualize This and Data Points books that guide readers through the process of turning raw data into compelling visual stories. His expertise has been recognized and featured in numerous publications, including The New York Times, Wired, and The Guardian.
Related Episodes
Different Ways to Listen to the Show!
New Ways to Support the Show!
With more than 250 guests and 10 seasons of episodes, the PolicyViz Podcast is one of the longest-running data visualization podcasts around. You can support the show by downloading and listening, following the work of my guests, and sharing the show with your networks. If you’re interested in financially supporting the show, you can sign up for my Patreon platform, make a one-time payment via PayPal, or shop one of the show’s sponsors.
Transcript
00:00.40
jschwabish
Hey Nathan welcome to the podcast great to like actually talk to you in person.
00:05.29
Nathan
Um, yeah, thanks for having me. It’s It’s good to finally see you.
00:10.41
jschwabish
Yeah, many years of us sort of like running in in parallel and and and and working in the space I’m really excited I reached out this time because you’ve got a new edition of your of your first book right? visualize this coming out. Ah, any moment and wanted a touch base and and talk about that and flowing data which is come which is what this is a how long 23 years is that right in 2007 so we we’re not yeah we can’t do the math. We’re not, we’re not goingnna try to math. But um.
00:37.23
Nathan
Um, since 2007 so yeah yeah some yeah yeah.
00:45.40
jschwabish
So so um I thought we could start with the book. So what? what’s the what’s the new addition. What what are people? What can what can folks expect.
00:57.42
Nathan
Um, and yeah I have the the second edition of visualize this is coming in June this year um it’s an updated version of a book of visualize this the first edition which I wrote it was published in 2011 but I wrote it in 2000
01:03.94
jschwabish
Um.
01:14.23
jschwabish
Is it.
01:14.94
Nathan
Nine Ten I guess and um I be I so I so I started updating the second edition or I started updating for the second edition as sort of like a very simple update. It’s but it then it started kept going.
01:27.46
jschwabish
E.
01:32.84
Nathan
Because I hadn’t written a book in a very long time and um I basically ended up changing all the examples and rewriting almost all the copy. So it’s ah it’s ah it’s basically a brand new book. But like the root of the my process the head.
01:35.66
jschwabish
Yeah.
01:51.83
Nathan
Was developing in 2011 um is is kind of is still there in the second edition.
01:53.55
jschwabish
E.
02:00.50
jschwabish
So does it is it following the same structure of the first book just ah updated for you know changes over in the field and and changes presumably in your perspective over the last decade or so.
02:15.18
Nathan
Yeah, the the structure is similar um the chapters some of the chapters have changed and I’ve rearranged things um that I feel um is more is better for learning how to visualize data. So the first edition. Um I kind of.
02:19.66
jschwabish
E.
02:33.20
Nathan
I wrote it when I was in grad school and I had just finished my my master’s and I was headed towards a ph d in stat and so the first edition was written as like a way of exploring data sets and focusing on the data and then going to the visualization versus.
02:38.37
jschwabish
E.
02:49.90
jschwabish
It.
02:52.84
Nathan
Um, focusing on like the design parts and the methods and the chart types so is framed is is with data types and so the second edition is still framed with data types and asking questions about data and answering questions.
02:57.30
jschwabish
You.
03:10.83
Nathan
And then sort of like iterating and exploring and then focusing on audience and designing for communication.
03:16.10
jschwabish
Right? And and because you come from like a math statistics background. How do you?? How do you think about writing the design aspects of stuff like that I always always coming from like an econ background I always feel a little odd writing about. Design when like you know I don’t have that training just you know, just kind of learning by doing.
03:42.22
Nathan
Um I Guess so like we go back if I go all the way back to when I started learning about visualization. It was um, it was about reports. It was like I was analyzing the data and I just needed to kind of make really quick charts.
03:55.67
jschwabish
Um, yeah, yeah.
03:57.43
Nathan
Put it in a report and it sort of like a second thought like an afterthought. Um, but then um I started working with personal data collection which was my dissertation topic and it was more like people were collecting data about themselves as a diary aspect to it.
04:12.22
jschwabish
Easy.
04:16.80
Nathan
And it was not so much about like analyzing yourself and trying to improve yourself in some way. It was more like collect data about yourself like you would write in the diary and then go back. You know five ten years later and see yourself in that light in that way. Um, and so I designed visualization in that way.
04:27.98
jschwabish
E.
04:36.40
Nathan
That was more like it was more contextual and it was more about like a queue to remember things that that happened later on so it was like a supplement to everything else in your life and then from there it went um.
04:39.91
jschwabish
Um, yeah.
04:47.35
jschwabish
Right.
04:53.87
Nathan
I ended up at the New York Times for a summer internship and then so that kind of like pushed me along for just it was it was only like six weeks but it kind of like spurred my thinking along and that has sort of influenced me.
05:02.88
jschwabish
Yeah, no.
05:12.10
jschwabish
Right? right.
05:12.16
Nathan
Pretty much like throughout my career and so like communicating to a very wide audience who doesn’t work with data um and taking sis will concepts like distributions and you know quartiles or whatever and trying to explain that to people who don’t work with who don’t analyze data.
05:24.49
jschwabish
E. Right.
05:31.75
Nathan
And so that kind of just like balloon from there and I ended up doing it for you know, 20 years
05:37.91
jschwabish
Yeah, um, do you have? Well I guess recommendations in the book or or in your head on how to more effectively communicate the uncertainty piece and the distributions piece I Know it’s like a big challenge for people. Ah, because like you said I mean you know you say quartiles or quantiles to people. Lots of people’s eyes glaze over like and I know you do a lot of this on the ah on your site and in your blogs and I want to I want to come to.
06:02.72
Nathan
Um.
06:07.60
jschwabish
Ah, some of the differences between like the static work you do and the interactive stuff you do, but just generally speaking like how do you think about communicating uncertainty to folks who you know may not have a stats background.
06:18.58
Nathan
Um, yeah I think um, yeah so I always remember when I was I was in Vegas with some friends and um as a stat student I I learned everything I could about blackjack and the probabilities.
06:25.67
jschwabish
Is it.
06:33.64
jschwabish
Easy.
06:35.98
Nathan
And then so I learned like the perfect can and when you would um when you would hit when you would stay and you know, um, and so I would tell them like so hit stay at certain points and then um, that wouldn’t always be like a winner.
06:41.12
jschwabish
Right.
06:53.83
jschwabish
Yeah, yeah, yeah, yeah.
06:55.55
Nathan
Like they would sometimes lose and they would get upset. Why did you tell me to do that when you know you you claim that the odds were in my favor if I did this so it was like this very simple so ah, uncertainty problem because.
07:05.91
jschwabish
You’re right? yeah.
07:11.70
Nathan
They looked at kind of like the individual data points but they had to you know extrapolate for like hundreds or thousands of of hands and so yeah, you always have to think about it long term. Um, so always I always kind of like think back to that and I guess people think.
07:17.20
jschwabish
Um, yeah.
07:31.25
Nathan
They often think about the individual data points like the ah current individual currences. But it’s harder to think long-term so I kind of link on to that part where to show the individual data point so that they can easily relate they can find themselves in the data in some way and then extrapolate.
07:43.23
jschwabish
Is it.
07:50.61
Nathan
That in some way So even with demographic data. You could someone can zoom in onto their about themselves about their own groups and then they can sort of go from there and extrapolate and look at the whole entire population but they can always have that they have the anchor of themselves.
08:03.92
jschwabish
Right.
08:10.60
Nathan
In that data. So it’s like it goes back to my personal data collection stuff where it’s like you have that individual point that diary entry thinking about yourself and then seeing how that relates and how you relate to the rest of the world.
08:14.64
jschwabish
Um, yeah.
08:25.59
jschwabish
Right? Um, does the are do you dive into like in does the new book dive into I mean that’s pretty ah, that’s a pretty nuanced date of his challenge does the book dive into those. Sorts of nuanced detailed aspects or it like how I guess the question is how do you view? this book is it for a beginner in Datavis is it for someone who’s been in for a while should they get the first edition and then the second edition do they roll in that sort of fashion.
08:59.75
Nathan
Ah, they would don’t get the first edition get the second edition. Definitely um, no, don’t but yet what? yeah so the second edition will be much more useful than the first because just because the tools are different.
09:06.94
jschwabish
Ah, but by 2 of them.
09:11.66
jschwabish
Yeah.
09:17.71
Nathan
The design is different like when I wrote the first edition we didn’t have to think about mobile flashed it. Flash still existed. Um, it was a time where like visualization on the web was kind of figuring its way out.
09:24.90
jschwabish
Um, yeah.
09:32.36
jschwabish
Yeah.
09:35.70
Nathan
And it was sort of like this translation of like powerpoint and so it was people coming from powerpoint and then making graphics for the web so it was like these really big infographics or long slide decks or slideshows.
09:43.59
jschwabish
Right.
09:52.55
Nathan
But now it’s sort of like you have I mean you have phones and that is a very really big chunk of what you have to do and which you have to figure out so I talk about designing for mobile. Um.
09:58.82
jschwabish
Great. Yeah.
10:08.97
Nathan
And thinking about your audience and designing for the data that you have and the tools that you have and it’s very much about, um, it’s for people who are figuring out like what they want to visualize how they want to visualize it and and what tools they’re going to use I try not to.
10:12.78
jschwabish
The.
10:26.91
Nathan
Focus on any specific tool I have so there’s an introduction to like a very wide variety of tools with concrete examples of using those tools because it goes it goes back to like when I learned visualization. It was a very informal process I never took a class and I sort of I’ve.
10:34.88
jschwabish
E.
10:44.27
jschwabish
Yeah.
10:46.89
Nathan
Picked it up as I went along and in my experience. It’s been that you you know you try to learn as many tools as you can and that makes you um your skills become extremely flexible in what you can do and with the data that you can work with and the kind of like the.
10:59.20
jschwabish
Is.
11:05.92
Nathan
The screen size that you can work in.
11:07.69
jschwabish
Yeah, yeah, so there there seems to be ah in the data is kind of book literature library and evolution in the books from the.
11:21.49
jschwabish
You know intro tufty few my books would fall into this like kind of like this intro like how to ah you know, sort of like setting the setting the the baseline and then there’s there’s newer books that I think are are. Are sort of evolving with the field like Jen Christensen’s book on science graphics and fidget settler’s book on um, functional aesthetics and those sorts of books. Do you are you trying to position like does does do you view this book as like the intro guide or is it within this. This sort of evolution of the field of we’re moving beyond like these quote unquote rules into the practical and how to push all those boundaries.
12:07.78
Nathan
Um, it’s it’s I would say it’s for people who know what visualization is but they don’t know how to do it and so they like there’s it. So when I wrote the the first edition it was very much like.
12:12.64
jschwabish
E e.
12:25.72
Nathan
Ah, handbook written to myself because um, so it goes back to that New York Times thing and I had um before that it was all about charts for analysis and charts for exploration I had done very little for charts for presentation and communication.
12:31.15
jschwabish
Um, yeah.
12:44.72
jschwabish
Um, yeah.
12:45.10
Nathan
Um, so I to try to prepare like in retrospect is totally worthless. But I I tried to I read as many design books as I could and like with the established design visualization books as I could and when I got.
12:53.54
jschwabish
Um, yeah.
13:04.35
Nathan
Ah, when I got to the New York Times I was I guess I was surprised that no one even I don’t know I think I had it in my mind that they like kind of pontificated about like the design aspects of like charts because there’s the New York Times is like they’re the best.
13:17.70
jschwabish
Yeah, yeah, yeah.
13:23.15
Nathan
Um, so like they’re definitely thinking about all the colors and like all the design aspects and their audience are thinking about all these Ti things and they you know si tea at it at a large table and and like point out all the the mistakes are it was like we just got to make this. We just got make it like.
13:34.65
jschwabish
Right? but. Um, yeah, great.
13:42.40
Nathan
What do you think is best and then try to and then and just go from there and then so there was like a bunch of tradeoffs of the technology that was involved and the people who are available the data that was available and so you kind of like take into all those tradeoffs into account and then you make something that is useful as useful as you can.
13:55.29
jschwabish
Um, yeah, right.
14:01.47
Nathan
And so like visualize this was trying to get past the all that big cluster of design books that I read that ended up not helping me I mean they helped me in my like original like kind of base thinking. Um, but then I like I just needed to actually do something like how do I use illustrator.
14:06.91
jschwabish
Right.
14:13.60
jschwabish
Um, right.
14:18.73
jschwabish
Yeah.
14:20.46
Nathan
Um, how do I use art in the context of communicating data and so I learned all of that like while I was there and then I’ve learned that over the years and so visualize this is very much geared towards this is like step by step how you do it.
14:23.93
jschwabish
Is it.
14:38.90
Nathan
I’m not waving my hands at all like I’m not like brushing over things like this is here’s a code where here’s the Files here’s the data and here’s what I did and you can do what you want to do and next step take your own data and do it. You know use similar things.
14:38.19
jschwabish
Um, yeah, yeah.
14:51.13
jschwabish
And do it. Yeah right now I remember for a while you were writing code in R to create the graphics and then would pipe them over to illustrator to add annotation and then clean that up. Do you still do that or has r. And I know you do D 3 stuff and and and lots of other things on the on the on the flowing data site but has r gotten far enough now where you don’t need to take that extra step over to illustrator as much.
15:22.10
Nathan
For me. No no yeah I think for a lot of lot of people I understand that you know it’s It’s like really really efficient to stick stick within the software and our has.
15:28.27
jschwabish
Ah, yeah.
15:36.64
jschwabish
Um, yeah.
15:40.81
Nathan
Is flexible enough where you can. You can definitely reproduce any graphic that you want um the advantaged so I use R and then illustrator still the advantage is that you don’t always know like how it’s going to look and you want to be like clicking and dragging to see.
15:43.63
jschwabish
Um, right? um.
15:53.59
jschwabish
In.
15:59.60
Nathan
You know what? a color looks like what a labels look like how it’s going to fit on the page how it fits in structure. Um, and so illustrator or any illustration software is it just lets me like see exactly what is going to look like and then be able to export it in different formats for for the web.
16:03.70
jschwabish
The.
16:13.70
jschwabish
Right? right? And and what about the work that you do um for interactive. So ah for your interactive pieces. Do you. And are you working are you for those for that for that for that task that you just mentioned of playing with color and moving things around. Are you playing around with static versions in in Illustrator or is you doing are you doing all that in the code and doing it all in the browser.
16:42.84
Nathan
It’s um, so my the now I always start with analysis and exploration because I usually don’t know what I exactly am going to visualize or how I’m going to do it. So for that R is my thinking language I Just pull the data into R and then I explore and look for.
16:45.97
jschwabish
Yeah.
17:00.67
Nathan
Like structures and visualization methods that would work and if I land on something then and I think that interaction or animation is going to make it easier to understand than I then I bring it I format the data clean it processes it.
17:03.99
jschwabish
Um.
17:13.81
jschwabish
Um.
17:19.80
Nathan
And then it code goes to javascript css and Html and then after that it’s all it’s all browser work. So I’m like doing trial and error. But in between the editor in the browser.
17:23.51
jschwabish
Um, okay.
17:29.72
jschwabish
Right? And do you think I mean I’m always curious about this because I do not code in javascript I’m but I’m always curious about this kind of the difference here like is there a reason because I could imagine creating like that first view in in the browser. And I can imagine you know, making that as an image and going into illustrator to sort of play around with annotation layers and and different pieces but sounds like you’re just working in the browser and so like what is the how do you think about the difference in your process between a static version where you have like 2 tools versus. Ah, javascript and interactive piece where you’re just kind of working in that single tool.
18:15.27
Nathan
Um I think it is um for me the interaction part the good coming into D Three is always is never my the first step. It’s always way on the tail end of my process and so it’s always in the language.
18:19.76
jschwabish
E. Um, yeah, yeah.
18:33.50
Nathan
For me is R is to so I I Just like don’t have to think about the syntax or structure or anything because I’ve used R for so long. Um, and so if I have a question I answer it I do it real quick and I can you know iterate in R you know for hours or days or whatever and then I’ll spend.
18:39.35
jschwabish
Right.
18:52.39
Nathan
Ah, much smaller chunk of time in javascript and coding it because even with even though I’ve been doing it for a while. There’s I still have to like look at things up with Javascript with web with web stuff because it’s always changing. Um and it’s it’s also just not my primary.
19:06.72
jschwabish
Um, yeah.
19:12.10
Nathan
Like analysis language and as some people use are able to analyze data in Javascript and so more power to them. But so it’s always like I will have a very concrete thing in my head before I head over to the web stuff.
19:13.10
jschwabish
Right.
19:19.40
jschwabish
Yeah, yeah, yeah.
19:29.35
jschwabish
Gotcha gotcha. Okay, um, so like like we talked about the beginning you’ve you’ve been running the flowing data site since since 2017 and the site has grown in some ways but sort of still has kind of a core set of things that you produce there’s like the daily. Email with an example that that that’s sometimes ah, mostly a data visualization. But other times it’s just like interesting stories that are connected to the to the field and then you’ve got the um the membership piece where you could take tutorials and d 3 and excel and R. Um, and then you get your own datavis projects and so I guess there’s kind of 2 questions for you like you’ve kind of like kept this kind of just rolling ahead and I’m curious about how you keep it fresh for yourself after after all these years and also like which. Part of the of the of your work is your most favorite thing to do out of all of those.
20:31.66
Nathan
Um, yeah, it’s um I yeah I I get this question I’ve gotten this question a few times it’s because it’s been so many years and I it’s for some reason I find things to do um.
20:37.23
jschwabish
Um, yeah.
20:47.99
Nathan
I think it like I have no plan for it because flowing data started as just sort of like a way to share things with my classmates so I was at ucla um I had to move to Buffalo after my master but I was still.
20:50.59
jschwabish
Um, yeah.
20:54.24
jschwabish
E.
21:03.17
Nathan
Working on my Ph D So I used flowing data which was a domain name that I got ah that I had to buy that was free because I had to use the hosting and then so I started sharing things on flowing data and sending it to my classmates and then sort of kind of went there. Um so I feel like.
21:15.71
jschwabish
E.
21:21.80
Nathan
If you go back to like the very beginning of flowing data and look at the early posts which is it’s kind of embarrassing for me, but it’s me like trying to learn and figuring out what visualization I think like there’s probably something some post that I wrote like what is visualization because I probably didn’t even know what? ah.
21:30.56
jschwabish
Ah, right.
21:40.70
Nathan
But I actually did not know what visualization is so it’s like me working things out in my head and then it kind of and now I make things and then sort of it’s suddenly kind of like caught on and other people found it. Um, and so just like throughout grad school I worked on.
21:40.20
jschwabish
Right? right.
21:52.60
jschwabish
E.
21:59.13
Nathan
Was learning visualization so in the very beginning it was about like learning visualization. What I wanted to do with it and how it like applied to the rest of my work and then I graduated um I finished my Ph D and then by that time flowing data was um enough.
22:06.91
jschwabish
And.
22:17.76
Nathan
To be my my full time work. Um, so I just kept going with flowing data and that was part like part advertising part memberships at the time Actually I think it might have just been advertising I can remember Yeah I think.
22:18.10
jschwabish
Yeah.
22:27.39
jschwabish
E. As I say I don’t think the the membership started later as I recall right? yeah.
22:37.71
Nathan
Yeah I think I think membership started a few years later so it was just advertising at the time and that was when independent publishers were able to get advertisers and make people it was not just Facebook and and Google Adsense um, and so by like.
22:41.35
jschwabish
Um, yeah.
22:47.30
jschwabish
Um, right right? yeah.
22:57.55
Nathan
For me the the way that it just keeps going is that um, most most of it is based on my curiosities and so like I’ll see things or things happen to me or read something and then there’s always always a question that comes up and I.
23:06.10
jschwabish
Um, yeah.
23:11.46
jschwabish
E.
23:14.67
Nathan
And usually I can try to answer that with data and so the curiosity keeps it going about the data itself. Not so much like what novel visualization can I use or like what fun like how can I get more audience. How can I get more people to look at my stuff. It’s more like.
23:18.39
jschwabish
Right.
23:29.52
jschwabish
Yeah, right.
23:32.66
Nathan
Have this question I’m going to try to answer it and if I have the question then maybe you know some people usually have that question too Sometimes a lot of people have that had that question and you know sometimes a handful but that just keeps it going.
23:45.21
jschwabish
Yeah, and do you for for the pieces that you create. Um I’m trying to say the last there was one that you did I think there was like a beast one we did on marriage or fertility or something like that by occupation or something is recently. When you when you create those because you’re just you know trying to answer a question. Do you always take the code and write a tutorial on it or is that just for 4 pieces where you think there’s a gap or or you. Solved it in kind of an interesting way or do you just always say I’m just going to provide this tutorial because someone is going to want to build a bswarm chart in Javascript and I might as well just give it to my members.
24:27.45
Nathan
Um, I It’s not. It’s not for every ah every example, but if um like if it’s a chart type that I haven’t written the tutorial for yet and then I have not or I if there’s a way I can generalize like the very specific thing that I made.
24:29.73
jschwabish
Yeah.
24:35.28
jschwabish
E.
24:42.58
jschwabish
Yeah.
24:44.41
Nathan
And make it easier to use that people can apply their own data. That’s a big thing like I want people to be able to take their own data set and visualize use that method and adjust it and not just like copy mine like I provide my data and what I did but I want them to.
24:56.52
jschwabish
Um, right? yeah right.
25:03.38
Nathan
Know how to do it. So if if there’s a way to do that and it’s useful then I write a tutorial or a but guide for it or I’ll just if I’m talking about like a very specific process I use the the weekly newsletter of called the process and like just kind of walks.
25:10.13
jschwabish
Right.
25:20.93
jschwabish
Um, right.
25:22.43
Nathan
Through my thinking and what I was doing Sometimes it’s just like I toyed around with a new feature or like new interaction that I haven’t but used before and so I’ll write about that and kind of it’s It’s still me like.
25:33.71
jschwabish
Um, yeah.
25:38.89
Nathan
Figuring it out back in 2007 I’m like like what is this cider for what is this is it useful I don’t know. But yeah, yeah, so that’s like but at this point my my approach with like anything new is just like throw it out.
25:39.49
jschwabish
Right? right.
25:48.98
jschwabish
We’ll find out once it’s done and see if anybody uses it. Yeah.
25:57.89
Nathan
And see what happens because it’s like and how it converges has converged to something and you want it to like diverge to new stuff.
25:59.22
jschwabish
Um, yeah I mean it it.
26:06.50
jschwabish
Right? Well I mean it’s also interesting right? because because you have this sort of members piece where I kind of feel this I kind of feel this way with my with my newsletter I kind of feel like a little.. It’s a little bit of a sandbox. I can sort of play around and and test things out and get feedback without having to you know worry about the I may not worry, but you know worry about the getting it quote unquote right? or not um and just kind of try some things.
26:31.31
Nathan
Um.
26:38.68
jschwabish
Even within it’s a you know? whatever. However, many people it is but just but just kind of try and see what people think.
26:44.93
Nathan
Um, I mean there’s I like I I I worry about people who are new to visualization and like they’re excited about it and want and make stuff and then they put it out there because it’s.
26:54.98
jschwabish
Um, yeah, yeah.
27:01.17
Nathan
Like they’re excited about it and they’re they’re just learning and that but the internet and social media and and Reddit is is often is like really critical and sometimes mean about it and so I’m worried that that kind of criticism would detract.
27:15.80
jschwabish
Um, yeah.
27:21.15
Nathan
People who potentially could be very excited and you know that excitement could turn into be like very good skills and visualization become very good and you know make something that we’ve never seen before but because you know a bunch of like armchair experts like took them down and then it kind of like.
27:27.51
jschwabish
Yeah.
27:36.99
jschwabish
We’re fake.
27:40.19
Nathan
Like I don’t want to make this I don’t want to do this? Um so with like with flowing data just the the public. The non-member part that I’ve been doing since the beginning is trying to like highlight the good stuff and I tried to like it.
27:40.41
jschwabish
Yeah.
27:49.72
jschwabish
Um, yeah.
27:59.11
Nathan
I Try to stay back or like I resist trying to criticize things too much. But for the for the process. The members newsletter those are people who are already into visualization. They’re trying to learn it and so they’re like more fair minded I would I would say.
28:06.10
jschwabish
Um, right.
28:16.54
jschwabish
Um, yeah.
28:18.91
Nathan
Um, because they’re trying to learn. They know what it’s like to learn um and they’re making things too so I can be a little ah be more critical with that audience versus just you know putting it out there and having a thousand people disagree with you.
28:27.70
jschwabish
Right? right? Yeah,, but it also feels like especially in the process emails a lot of the stuff that you critique in those emails is not like I didn’t like this because of the color or it’s.. It’s not like a lot of the I don’t want to say subjective stuff because some of those design things are not really subjective but a lot of the critiques that I see you write about are more about misinformation stuff. That’s misleading as opposed to what someone could construe is your subjective opinion as opposed to. This person wrote this thing and like but the math is wrong, right? like so that that to me seems like a different animal and and a lot of ways.
29:09.30
Nathan
There.
29:16.53
Nathan
Um, yeah because I mean since flowing data has been around so much like people can go back and look at my stuff usually like you’re in a glass house. You should be throwing bricks. But um.
29:24.79
jschwabish
Um, you’re right right? Yeah right? yeah.
29:32.70
Nathan
And I remember I know what is like to like make things and it becomes popular and just have people dump on it so hard. Um, and like now that’s funny to me but I’m trying to imagine myself early on like what if I was um.
29:38.65
jschwabish
Um, yeah, yeah, yeah.
29:51.22
Nathan
What if I started flowing data like in my first year and then I did that and someone like all these people started dumping on me I don’t think flow and day would be around because I just wouldn’t want to do it.
29:56.56
jschwabish
Yeah, right? or right have you ever? You’ve had a bunch of things that get picked up by you know by by you know, big news organizations have you ever had anything picked up or you’ve published something that you thought was pretty innocuous like it’s on basketball data or it’s On. Marital rates and you get people dumping on you like has that ever happened where you’re just like what is the what is the problem.
30:22.42
Nathan
Um, um, yeah so I have I I Yes that has happened I have I have a series called Data Underload which is like a semi weekly chart series and it’s it’s.. It’s a little different now than when it started when it started. It was um, it was illustrative and not to be taken Seriously, it was like almost like a data comic. Um, so I would like in my mind make very obvious. Um.
30:48.91
jschwabish
Easy.
30:59.38
jschwabish
Yeah.
30:59.70
Nathan
Like insights with made up data and geometries. But then it might be misconstrued as reality because I didn’t It’s not it doesn’t look like a you know like a Sk Cd Comic or something like that. It looks like someone spent time analyzing data and making charts.
31:12.50
jschwabish
Um, yeah, yeah, great.
31:19.67
Nathan
So like I made um I made a chart with like sleep habits when you’re old and as you get older. So when you’re young, the sleep habits are like very structured.
31:28.17
jschwabish
Um, yeah.
31:36.88
Nathan
You become a teenager is that becomes like kind of all over the place and then when you get older, It’s sort of like kind of spotty. You wake up early and go to bed early and stuff like that. But it was I don’t remember the exact annotation that I use but it was it was totally made up. It was subjective by me. Um.
31:46.40
jschwabish
In.
31:53.40
jschwabish
Um, yeah.
31:56.86
Nathan
And so I get a lot of emails it I think I was on Reddit or dig or something and people are offended by it because I was making I made I made the mistake of using sarcasm on the internet and assuming that people would use the content.
31:58.67
jschwabish
Um, right.
32:07.21
jschwabish
Right.
32:16.48
jschwabish
Right? yeah.
32:16.84
Nathan
Take it into context but that but they didn’t so they get mad. But you so you like even even late many years later I took that same chart and I used actual data from the american time use survey like I recreated the chart that I made subjectively and use real data.
32:26.93
jschwabish
No, right? It did it match. Did it match up. Yeah, right.
32:34.83
Nathan
But people still got mad about it is so weird but it did match. It is like is very close like but it was more granular because I had actual data to look at and um and so like a lot of the things that people get mad about is when.
32:45.37
jschwabish
Right.
32:51.94
Nathan
You show a pattern It doesn’t match their experience and so it seems like it’s wrong which I understand but ah it goes back to like that blackjack problem where you have the single thing like that’s wrong, but you know Overall we’re looking at the whole population looking at everybody.
32:52.52
jschwabish
Yeah, yeah, break.
33:03.00
jschwabish
Um, yeah.
33:10.29
jschwabish
Yeah, right? ah.
33:10.46
Nathan
And so it’s a little different for everybody and so I think that that thing is like if you link on to the individual thing again that makes the everybody part much more understandable.
33:19.25
jschwabish
Right? right? it it was funny because you had a couple of of posts and and emails in the process where where you were using the the time you survey and I was like oh Nathan’s like gotten like neck deep into the time you survey and he’s like I’m just gonna just gonna. Ring as much stuff out of this as as possible and so I saw like you know there was like this series. There was like this series of posts I wish I made a graph of like Nathan’s use of a particular data set because like that one sort of like got really big and then sort of tailed off again. So um, you did that you had that a few years ago I think with like ah. Maybe the l la h d or something from the Bls. There was like a lot of like employment stuff that that that was showing up so I I kind of do appreciate like the evolution of seeing your use of different data sets sort of show up over over time. Um.
34:08.60
Nathan
Um I I love that data set that is that is my favorite data data set. Yeah.
34:14.44
jschwabish
The time use one? Yeah um, so we’ve we’ve covered a lot of ground so we started way back in in in grad school and o 7
34:24.94
jschwabish
The the visualize this volume one and then the new edition and and what you’ve been doing at the site and I’m just curious to to finish up here. What? um, looking forward into the future data is whatever that’s going to be like is there something that excites you Ah as you see what people are doing or what you see that. You know might might come down the down the road.
34:51.62
Nathan
Um, well I Ah I guess because I when I think of visualization in general. It’s sort of like this. It’s this abstraction of data and it’s like very, it’s sort of like this is disconnected from the real world in that.
35:09.10
Nathan
Um, the data is ah also data itself is abstract the visualization is abstraction of the data and so people have to like get from the geometries to the math and back to Reality. So I think as the computers have gotten better. The. Tech has got better and people understand data more can kind of like shorten that path between reality and the abstraction and like um, making visualization less of like that thing that’s on a report that thing that is like on the side or is a separate.
35:31.92
jschwabish
If isn’t.
35:48.32
Nathan
Part of an article or anything. It’s sort of the data is just there the chart or the visualization is just there and people ah can understand it in some way that has to do with Well it has to do with people understanding data and.
35:48.87
jschwabish
Yeah, yeah.
36:01.77
jschwabish
Is.
36:06.89
Nathan
So I’m hoping that I I mean there are different like mediums that are coming out to kind of shorten that path like virtual reality and and augmented reality. Um, so I I use.
36:20.12
jschwabish
Um, the the.
36:23.99
Nathan
I use virtual reality I use Vr ah vr hesit for fun I wonder if like data could be brought in that that way where you’re interacting and more in a more tangible way. Um, and even just kind of like.
36:27.71
jschwabish
Um, the.
36:31.76
jschwabish
Um, right.
36:39.59
Nathan
Instead of having charts photos videos sort of like their own separate thing. They’re sort of more intertwined so data understanding the data and the context of the data is sort of like this kind of like United thing versus sort of separate things. Yeah, and so I think we’re kind of like getting towards that but it’s.
36:49.14
jschwabish
Right? Yeah, separate separate things. Yeah, yeah, yeah I mean it is still surprising to me how many times you’ll read some sort of report.
36:57.30
Nathan
Store Pretty long way to go.
37:05.30
jschwabish
And they make a pretty clear argument in the report. But then when you look at the graph The graph is just like this description of you know this is the data in the graph and doesn’t tell you anything about the argument that they’re making So I yeah it’s still a battle to be fought I think.
37:17.17
Nathan
Um, yeah, yeah I think the like the writer like often people who are reporting the data don’t Also don’t understand the data to its full extent. It has to do with like the uncertainty of it and the probabilities and.
37:30.79
jschwabish
Um, easy.
37:35.88
Nathan
Kind of um going beyond like the reported Median and mean that’s sort of a big focus if people can understand that it’s very fuzzy and there is like it’s It’s usually not certain almost everything.
37:38.20
jschwabish
E.
37:52.74
jschwabish
Right? yeah.
37:53.31
Nathan
And Data is not certain so if they can like figure out how to communicate that in some way or to understand it and have just have a general sense that this may or not may or may not be the actual thing it would be I think it would be more useful.
38:05.99
jschwabish
You know? yeah for sure. Well Nathan congrats on the new book looking forward to to seeing it in June and thanks, thanks for coming on the show as always it was. It was great cha with you.
38:17.92
Nathan
Yeah, thank thanks a lot for having me This is good to be on.
