Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Charting Success: Nick Desbarats’ Practical Approach to Data Visualization  image

Charting Success: Nick Desbarats’ Practical Approach to Data Visualization

S10 E259 · The PolicyViz Podcast
Avatar
812 Plays7 months ago

On this week’s episode of the podcast, I speak to author and teacher Nick Desbarats about his new book, Practical Charts: The Essential Guide to Creating Clear, Compelling Charts for Reports and Presentations.  Our conversation covers choosing appropriate chart types, emphasizing simplicity and clarity, and understanding when to use different formats. Nick aims to challenge dogmatic views on charts, such as the use of pie charts, and stresses the importance of catering to the audience’s familiarity with graph types. Our chat also includes insights on transitioning to online teaching during the pandemic, the distinction between clear graphs and dashboards, and the significance of mastering fundamentals in data visualization for beginners and intermediates. If you’re familiar with Stephen Few’s work, you’re also bound to find some behind-the-scenes gems in this week’s episode.

Topics Discussed

  • Choosing the Right Chart. Nick kicks off our conversation with an essential primer on selecting the appropriate chart types for different datasets. His focus is on simplicity and clarity, ensuring that the chosen chart communicates the intended message as effectively as possible.
  • Challenging Chart Dogmas. Prepare to have your preconceptions challenged as Nick takes on the controversial stance on pie charts and other commonly debated graph types. It’s all about breaking the mold and understanding why certain charts work better for specific audiences.
  • Catering to Audience Familiarity. A significant portion of our chat is dedicated to the importance of tailoring chart choices to the audience’s level of comfort and familiarity with different types of graphs. This segment is crucial for anyone looking to maximize the impact of their data presentations.

➡️ Check out more links, notes, transcript, and more at the PolicyViz website.

Sponsor: Whisper Transcribe

Whisper Transcribe is an absolute game changer in the world of transcription and AI-generated content creation! Get an accurate transcript from your audio and video files and automatically generate content directly from it.

Check out Whisper Transcribe

Recommended
Transcript

Introduction to Policy Diz Podcast

00:00:12
Speaker
Welcome back to the Policy Diz Podcast. I'm your host, John Schwabish. Happy Spring, everybody. I hope you are well. I hope your allergies aren't too bad. Here in Northern Virginia, we've got what we call the Great Pollening. It's this green sort of dusty pollen over everything, which is making our eyes water and our noses itch. But we power on making our way towards summer.

Introducing Nick Deborah and 'Practical Charts'

00:00:35
Speaker
On this week's episode of the show, I welcome Nick Deborah, author of the new book, Practical Charts, the essential guide to creating clear, compelling charts for reports and presentations. Now, if you are new to the field of data visualization, this is going to be a great episode for you.
00:00:51
Speaker
because you're going to learn some of the key features of creating better charts and graphs from Nick's work and from his book Practical Charts. He's going to talk about ways in which to choose your different chart types and how there is or may not be rules, guidelines, strategies for creating better charts in your work.

Learning from Steven Few

00:01:11
Speaker
If you are a more experienced data visualization creator or have been in the field for a while, you're going to like this episode because Nick used to work with Steven Phew. Now, if you don't know that name, Steven Phew, Steven was one of what we might consider sort of the, one of the modern, big people in the field of data visualization. Wrote a great book, show me the numbers and had a blog and some other great work he has since retired. But.
00:01:37
Speaker
he worked with Nick on Nick's teaching and instruction. So you're going to get a little bit of behind the scenes action on how Nick worked with Steve and developed his course, developed his instruction, and ultimately wrote this book to go along with that.

Audience Engagement Request

00:01:53
Speaker
So in this episode, you're going to hear about how Nick approaches data visualization, the thinking behind choosing different chart types when some are more or less appropriate. We're also going to talk about the distinction between making graphs simple versus making graphs clear. And we also finish up the episode talking a little bit about the distinctions between dashboards and more static graphs.
00:02:15
Speaker
Now, before I put you off onto this episode of the show in the interview, just a quick request. If you could take a moment out of your day to rate a review this podcast on your favorite podcast provider, I'd really appreciate it.

Nick's Journey into Data Visualization

00:02:29
Speaker
Those ratings, those reviews go a long way towards attracting more guests to join me on the show so that we can help more people be better and more effective data communicators. So just
00:02:41
Speaker
Click on over to iTunes, Spotify, wherever you get your podcasts, click those five stars, write a quick little review, won't take you very long. I'd really appreciate it. It helps the show be that much better and reach that many more people. So on to my discussion of practical charts with author and instructor and teacher, Nick Debra.
00:03:04
Speaker
Hey, Nick, great to meet you. Very excited. Likewise. Welcome to the show. I feel like we've been running parallel circles for a decade, maybe? Probably. Probably something like that, yeah. Yeah, long overdue introduction conversation. Maybe I just aged the both of us a little too much, too early. Yeah, that's true. We've only been on the internet for a year. Right. Yeah, right, right. Just found out about this whole data thing. Yeah, right.
00:03:32
Speaker
Um, so you have, uh, you have this new book out practical charts. Got it right here. Oh, you've got it behind you too, which is a good CNN, CNN, MSNBC move. Yeah. So practical charts and then more practical charts, excited to chat with you about the book and your work. But maybe we start with a little bit of background, uh, how you got into database and then, and then how you got into deciding like needed to write a book about the work

Teaching and Course Development

00:03:58
Speaker
that you do.
00:03:58
Speaker
Sure. Yeah. So way back 50,000 years ago, I was a software developer and got a little bored of that after a few years, moved around different areas of software organizations, doing some product management, and then they ended up doing a bit of marketing.
00:04:13
Speaker
business development, bit of sales. And then probably maybe I guess about 15 years ago, I just kind of stumbled on the field of cognitive psychology, which I just found absolutely fascinating, you know, cognitive biases, judgment, decision making.
00:04:30
Speaker
psychology of perception. And I just started really inhaling a lot of books in that field, which I consider to be just kind of a sideline interest. It wasn't kind of my day job, right? I was still basically doing software during the day. And then I was working on a product that had a lot of data visualizations in it. And like a lot of people just kind of muddling my way through and you know, doing what I thought kind of looked good or was probably the right thing to do.
00:04:57
Speaker
But then I figured I should probably get a bit of formal training here. And so I attended a workshop by Steven Few, who probably a number of your listeners are going to know. But Steven is a very well-known person in the field, did a lot of groundbreaking work. And that workshop just blew my mind, really.
00:05:19
Speaker
It was kind of the intersection of my day job and my sideline interest in psychology, right? It was this mix of data and technology and psychology. And so I was, yeah, I was absolutely fascinated, ended up kind of staying in touch with Steve after the workshop.
00:05:39
Speaker
And then a few months later, the conversation kind of evolved. I pitched him on maybe teaching his workshops. And he had had people who had kind of approached him before about teaching his workshops.

Adapting to Online Teaching

00:05:52
Speaker
But really, you know, if you've ever read any of his books or been to any of his workshops, you know that it is actually more like 80 percent psychology.
00:06:01
Speaker
And so most of the people who had approached him really only had technical backgrounds. They didn't really have a psychology background, but because of my sideline interest that I'd had for a number of years by that point, you know, he agreed and then trained me to teach his workshops, gave me a big sack of books to read. And yeah, and then the very first workshop that I taught was at NASA.
00:06:25
Speaker
Oh, wow. Yeah. So it was really kind of jumped into the deep end. Yeah. That was in, I guess, about 2014. Okay. And then saw, so I tossed these workshops for all over the world, for over a dozen countries until 2019, when, much to my surprise, Steve announced that he was going to retire.
00:06:46
Speaker
which I didn't know was in his plans. And so he actually encouraged me at that point and said, oh, you know, you can still continue teaching my courses, but if you wanted to develop your own, you know, maybe eventually write some books, then, you know, encourage me to do that. Right. And so I started working on a dashboard course first called Practical Dashboards.
00:07:08
Speaker
Uh, and then, uh, database fundamentals course, uh, which was the practical charts course launched them straight into the pandemic. Uh, like everybody had pivot. I'd been asked to teach online, but before, but I, I'd always refused because I was like, oh, it's going to be such a terrible user experience. You know, nobody wants to watch you talking head on a screen for hours on end, but.
00:07:30
Speaker
We all had to adapt the world. Yeah. Yeah. And so, uh, so yeah, I, I, you know, did a bunch of research in terms of how to teach online effectively, rejigged the courses. Cause you can't just, as you're probably aware,

Specific Guidelines vs. General Advice

00:07:42
Speaker
you can't just transport something that was designed for.
00:07:45
Speaker
in person and use it as is online. And thankfully it went really well. The courses are and were quite popular online. And now I've started to do a bit in person, but a lot of organizations still want online because they hired so many remote workers.
00:08:06
Speaker
you know, just to get everybody in the same room would be a big deal. I'm hoping to do more in person, but it's still mostly kind of online at this point. What is the biggest difference between Steve's course, kind of the original Steve's course versus what you teach now? Well, I mean, you know, I should qualify anything I say that
00:08:28
Speaker
you know, everything I do would not be possible without the enormous number of things that I learned from Steve. Not just about data visualization, but about teaching and pedagogy and retention and comprehension. But I guess, you know, after teaching his courses many times, you know, I started to see
00:08:49
Speaker
maybe some opportunities to provide more specific guidance than what I saw, not just in Steve's courses, but in fact, in other books and courses that I was aware of. I felt that the advice, especially around chart type selection, tended to boil down a lot of cases to
00:09:11
Speaker
use your judgment or do what looks right. But I knew in teaching people who often had actually very little background in data visualization that wasn't helpful. They're like, I just don't have that experience. I haven't developed those intuitions yet. And so just through a lot of iteration, I started to develop guidelines around
00:09:35
Speaker
you know, mostly chart type selection, but also kind of other design choices as well that ended up being what I felt were kind of more sort of usefully specific. Instead of like things that were more kind of vague in terms of like, Oh, you know, you got to break down the total.

Flexibility in Data Visualization Guidelines

00:09:52
Speaker
Well, you could use a bar chart or pie chart or a stock bar chart or.
00:09:56
Speaker
a regular bar chart or a Pareto chart or a waterfall chart, use your judgment. Use the one that you think kind of works best. It's like, actually, no, there are specific circumstances and conditions under which it does and doesn't make sense to use each of those chart types. And they're not interchangeable. You're well aware. There's a difference if you're showing the same data as a regular bar chart or a stack bar chart, for example.
00:10:25
Speaker
different kinds of insights are going to be more or less visible. And I just felt that it was possible to sort of get more formal about that, codify it a bit without being dogmatic. I don't use the term rules. I call them guidelines because this is the way I do it. And if you want to do it another way, that's fine. I'm definitely sort of okay with that.
00:10:52
Speaker
And in fact, if somebody like yourself who has a lot of experience is going to probably deviate from those guidelines every so often, and that's fine. Really, I consider them to be more like almost like training wheels, especially for people who are starting out, although even people who have quite a bit of experience have said that they also find that to be useful.
00:11:14
Speaker
But it basically kind of, you know, it starts you out with a good base, right? Like if you want to know kind of when to break the rules, the first step is kind of to know what the rules are in the first place, except I hate that term. I call them guidelines. And people have really responded to that as something to kind of latch onto in a field where there tends to be a lot of, that tends to be quite vague around a lot of these visions.

Critique of Box Plots

00:11:44
Speaker
Yeah, essentially, I wanted to ask you, because the book, the way you write it, it kind of dances around this rules, guidelines, strategies, preferences, I don't know, web of trouble when it comes to those words. But it's also interesting, and I don't want to belabor on talking about Steve too much. But I think a lot of people, I think, felt that Steve's writing, particularly on his blog, was pretty dogmatic about do this,
00:12:12
Speaker
don't do this, but what I'm hearing you say is, it's kind of like the old matrix, like you can kind of bend some of the rules and then you can start to break the rules. So like, how do you like kind of help people see the rule guideline and then just ignore all of it and, you know, go forward?
00:12:33
Speaker
Like I said, I don't consider that anything in my books are kind of unbreakable. Actually, it's not entirely true. There are a few points in the book where I say this is a mistake. If you show the data in this way, the risk that people are going to misinterpret the underlying data is so high that you should just never do this.
00:12:56
Speaker
But that is relatively rare, right? Yeah, like most of what I, I kind of advocate, I would definitely put in the, you know, the category of guidelines. It's just that I think, yeah, and I do have sort of an aversion to
00:13:12
Speaker
not just Steve, but anybody who is very kind of dogmatic about like, this is this is correct. This is incorrect. You know, it's like, well, you know, a rarely that simple. Right. Like there are often exceptions. And in fact, that's one of the reasons why I wrote the book is to kind of dive into that and say, OK, let's talk about those exceptions and those cases where, you know, a general principle maybe kind of doesn't apply.
00:13:41
Speaker
And so, you know, and this is one of the kind of things that I struggle a little bit with Tufti's books, Edward Tufti was another person that, you know, your readers may or may not be familiar with, but he's one of the godfathers of visualization that
00:13:57
Speaker
When I read his books, I'm often thinking, yes, this is usually a good idea, but not always. And those exceptions are, in fact, very important because those exceptions in certain cases can occur quite frequently.
00:14:13
Speaker
And so, you know, one of my goals was to really, like I said, to go beyond that in terms of like, let's actually flesh it out in terms of like, well, what are the exceptions? What are the specific circumstances under which, for example, it might actually make sense to use a pie chart. And, you know, where, you know, both Steve and Tufti would say, like, never use pie charts.

Challenging Misconceptions in Data Visualization

00:14:33
Speaker
Well, you know, pie charts are often misused. They're, you know, maybe a bit overused, but does that mean that you should never use them?
00:14:41
Speaker
Well, no, it's just that figuring out when to use a pie chart is actually surprisingly tricky. You know, like I have, you know, you've probably seen the book, I have these decision trees about which chart types to use and, you know, to show the breakdown of a total, for example. And, you know, there are like at least six major chart types for doing that, you know, stack bar, regular bar.
00:15:02
Speaker
Pareto chart, waterfall chart, and pie chart. And there are about eight considerations that you need to take into account when you're trying to figure out which is the best one for the situation at hand.
00:15:15
Speaker
And so there is a path that leads to pie charts. It's just that knowing what that path is, is not as straightforward as something that's as simple as like never use pie charts or always use pie charts to show the breakdown of a total, right? Like not, neither of those are good advice. It's unfortunately, it's just a little more complicated than that, but it is possible, I think to codify it because.
00:15:39
Speaker
you know, a lot of, you know, another sort of common thing that I hear, especially amongst even more experienced chart creators is, you know, they say, well, this kind of thing is just really, you know, impossible to codify. It's just, it's too complex, it's too nuanced, there's too many exceptions, you know, it's just something you kind of develop with experience and judgment over time.
00:16:03
Speaker
And so there I kind of pull back and go, well, you know, yes, it is complex, but it's not that complex. I think it is possible to actually codify it in a way which is, you know, relatively simple, but it's not as simple as like always use pie charts for the breakdown of a total or never use pie charts. That's too simple. And so so I kind of I think hopefully found a good kind of middle ground in terms of
00:16:29
Speaker
something that is not so simple that it's not going to be useful or it's going to give you bad design decisions really often. But it's also not this kind of nebulous cloud of experience and judgment and intuition either, because that's not helpful for especially for beginners.
00:16:47
Speaker
Right. Yeah. It's interesting because it seems to me when as I read through the book that the target reader, as you've mentioned, is kind of that somewhat beginner, maybe intermediate person creating data visualizations. But you also do have a chapter on motion and a chapter on interactivity, a chapter on animation. They're kind of, they're shorter chapters. And so is your avatar of the target reader
00:17:14
Speaker
that sort of beginner intermediate person creating static graphs and reports or on social media and not so much worried about that interactive dashboards or what things look like on mobile phones. Is that sort of like the core person you're thinking about?
00:17:32
Speaker
Yeah, I mean, in terms of if you think about it more kind of from a use case perspective, people are creating, you know, what I call everyday charts, right? These are charts for reports and presentations. It's not data art. They're not highly technical, scientific graphics, not like really
00:17:51
Speaker
you know, customize interactives like you'd find in the Washington Post or New York Times, not really my target reader. I think that people who are, you know, creating those types of charts would probably benefit from reading the book to maybe kind of get a little bit more solid on some of the fundamentals. But, you know, really, I only bring people as far as those kind of everyday charts.
00:18:15
Speaker
And a lot of people tend to think, oh, well, but you know, do you really even need to read a book to create these kind of simple everyday charts for reports and presentations? And yeah, you do. I mean, as you well know, like it's really easy to screw up a bar chart, you know, or maybe to use a bar chart when in fact you should have been using a line chart or a pie chart or something else. Right. Like these.
00:18:37
Speaker
These decisions actually require a surprising degree of skill, even if you're talking about these quote unquote, you know, simple charts. Yeah. Back to the rules part real quickly. Are there any, I'll use the word rules, although I, you know, there's definitely quotes around that, but do you have any rules that you view as ironclad?
00:18:57
Speaker
Like, for me, I think I can only think of one. But I'm curious if there are any that you have that are like all like all or, you know, never, ever do this or, you know, always do it this way. We'll have to know what your one rule is

Balancing Simplicity and Clarity

00:19:10
Speaker
now.
00:19:10
Speaker
Well, I want to see yours. Oh, okay. I'll give you mine. So I think, I think mine is the bar chart where people sort of break the bar. You know, they have like the outlier bar and they're like, well, I'm just going to like put that little jagged line and just going to shorten it. Because I think that I can't think of any exceptions where that just distorts.
00:19:32
Speaker
or perception of the data, right? And it's just arbitrary where you're going to cut it and how you're just going to adjust your axes. And I think a lot of people would say bar charts start at zero, but I think there are probably exceptions to that.
00:19:49
Speaker
you know, if you're gonna start, if you're gonna normalize your data at say one, well, so, you know, in that case, one is kind of equal to zero, right? But like, okay, so is one the right point? So, but I think that broken bar chart, what you talk about in the book, I think that might be for me, like the ironclad rule, which I rarely talk about in classes, which I guess I should, but it seems so obvious to me, like those most arbitrary things to me are the ones that I get nervous about, but yeah.
00:20:19
Speaker
Yeah. I mean, uh, I think if, if I do have any quote unquote rules, uh, where I just, and I like the way you described it. It's like, it's just, it's something where I just cannot think of a scenario where that's a good idea. Right. So it doesn't mean that they don't exist, but it means that like, I have not come across a good use case for it yet.
00:20:41
Speaker
Yeah. And so if you want to call that a rule, you know, sure. Sure. Yeah. But and so any of the kind of rules that I would have would be of the negative type. Right. There's never and always do this. Right. It's always, you know, never, never do this. Yeah.
00:20:59
Speaker
And that would, I would agree. Yeah. Like never break the scale quantum scale. Uh, I cannot think of a situation where that's going to be, uh, you know, a good idea. Um, what would be some other nevers? Uh, well, uh, a controversial one that I have is, uh, never use box plots. Oh, never. Okay. So I'm curious about that one.
00:21:20
Speaker
Yeah, so I wrote an article about this on Nightingale, the journal of the Data Visualization Society a year or two ago, and it just blew up. It was, you know, got hundreds of reactions and interactions. But I, you know, I used to teach bar charts, like, you know, bar charts were in Steve's courses, or no, bar charts, box plots. But I just, you know, after seeing the thousandth person kind of be like,
00:21:50
Speaker
And then eventually kind of like, oh, okay, I get it. It really sort of started to make me rethink like, okay, like there's just a huge cognitive hurdle associated with this chart type. When is it really doing something that alternative chart types can't do? When is it showing some kind of insight or answering some kind of question that you couldn't answer with a simpler chart type, like a strip plot, for example, which is much easier to understand.
00:22:16
Speaker
or even a histogram, which is a little more complicated, but still a lot easier to understand. People don't have to understand quartiles or mediums or anything like that. And I couldn't come up with any scenarios in which the insight that I was trying to communicate
00:22:35
Speaker
uh, could be communicated in box plots, but could not be communicated in simpler chart types. I looked and I, and when that article came out and went viral, I asked people, I was like, look, you know, prove me wrong. Like show me examples of data where you're showing the same data as a box plot, but then also the strip plot and histogram or stacked histograms in that case, because you need several histograms.
00:23:00
Speaker
And then tell me what you're seeing in the box plot that you're not seeing in the simpler chart types. I just didn't get any. So I vaguely remember that article, but let me ask, and then, and then maybe I'll, I'll give you an example, but where I think it could work. But when you say, uh, no box plots, are you again thinking about those, the way you, you mentioned earlier, like the everyday chart or because, because I can imagine for the everyday chart reader.
00:23:30
Speaker
you know reading the Washington Post or you know Twitter or something like that. I totally agree like people you know most people just don't understand or you know don't have experience reading what a percentile is or a quartile is and totally get that. But like in a scientific journal or an academic peer-reviewed journal you know that cognitive hurdle is probably not as high because those readers are probably familiar with those graph types.

Purposeful Charts Beyond Data Display

00:23:55
Speaker
I'm always fascinated for example when I like
00:23:57
Speaker
dive into the biology literature, which is very rare and very brief. But like I see these chart types and I'm like, I have no idea what you guys are showing, but like I'm not a I don't have a PhD in biology and have no idea what they're talking about. But I'm sure, you know, the editors and peer reviewers are like, oh, yeah, this is a blog chart. Right. So
00:24:18
Speaker
Yeah. Um, even in those cases, right? Like towards the end of the article, I kind of addressed that. Uh, and because even if you have like expert statisticians who have been looking at boxplots for years and years, like there's still, which, which I, you know, I fell into that category, by the way, I'm not an expert statistician, but like, I had been looking at boxplots for a long time. I knew exactly how to read them. I was very familiar with them.
00:24:42
Speaker
But then even with with me, when I would take the same data and show it as a strip plot or as, you know, stacked histograms, I was still able to read them much more quickly and easily. I often spotted trends, anomalies, patterns, gaps, clusters that just weren't visible at all in the box. You're just picking up the even with like expert audiences. I still made the argument that, no, I think there's even they are better off with, you know, these kind of
00:25:11
Speaker
I mean, I was going to say simpler chart types, but they're not just simpler. They're also more informative and less cognitively demanding, regardless of what your level of expertise actually is. And I made the same argument with connected scatter plots as well, and another article last year about that.
00:25:28
Speaker
Got a lot of attention, but I basically, you know, I subjected it to the same task. Like, can you take the same data and show it in this kind of very cognitively demanding chart type, but show it in simpler chart types? Like in that case, you know, the alternatives were just like two line charts that were stacked on top of one another or an indexed line chart. And then tell me what you can see in the Connector scatterplot that you can't see in these simpler chart types.
00:25:57
Speaker
And nobody could come up with anything, no use cases. And so I was like, okay, it's not that connected scatter plots or box plots or bad chart types, it's just that there are simpler alternatives, which are always able to say the same thing and in often cases more. And so why would you use those chart types?
00:26:18
Speaker
Well, I think and this was this is something I want. This is this is a good segue to what I wanted to talk about because you you start the book right at the beginning saying, you know, you've heard people say things. He's like, keep it simple and know your audience and tell a story. And then you sort of go into this.
00:26:33
Speaker
there's more to it than that. But this concept of simplicity, I hear people say that all the time and I don't love that word because I think what we're trying to do is clarify, not necessarily keep it simple, but to the connected scatterplot, I again think there's differences in your goal. I think back to like Hannah Fairfield's kind of like, not original scatterplot, but the kind of ones that made the connected scatterplot kind of famous is,
00:27:00
Speaker
She definitely could have created that as two line charts or an index chart. But again, if you are, you know, you're a Sunday afternoon New York Times reader, like I think in that case, at least, that was more about the experience, right, of reading this chart with the annotation to sort of bring you through this journey. Yeah, maybe two line charts or an index chart would have made that simpler.
00:27:25
Speaker
in the sort of sense that it's two line charts, we don't know how to read line charts, move on, but to engage a New York Times audience on a Sunday afternoon and bring them through that what I would call more of a story than what most people think of as stories. You know, in that case, I think it achieves a goal.
00:27:45
Speaker
Right, as opposed to if you're writing an academic peer review article or you're writing a report where you're like, here's this hypothesis and here's the evidence to support it. Yeah, the connected scatterplot I think more times than not doesn't really do that job.
00:28:01
Speaker
Yeah, I mean, that is definitely a very valid point, right? And in the beginning of the book, really the main kind of message of the introduction is I consider charts to be graphics for doing a job. And this is quite different than how most people think of it, right? And so if the job is just to essentially kind of grab attention,
00:28:24
Speaker
That's legitimate, you know? Yeah. Right. Like a lot of people say, Oh, no, no, no, that's not like proper data visualization. Yeah. No, no, no. Like sometimes that is, uh, that is the goal. You just have to be aware that if you're going to use some very unusual visualization or very cognitively demanding visualization to get attention, that there are going to be costs associated with that. Uh, you know, like in, in my connected scatterplot article, I actually talked about that New York times.
00:28:52
Speaker
You know, um, yeah. And I was like, you know, I, I would bet good money that the vast majority of readers did not understand that chart. They relied entirely on the annotations. Right. Right. Uh, or they misinterpreted it. They thought it was a standard line chart. And then they're coming out with a completely wrong understanding of the underlying data. And so yeah, it might've gotten a lot of attention, but at what cost? Right.
00:29:22
Speaker
So, yeah, I just, I guess to your point, there's, there's, we have different purposes, right? Like, yeah, you're just wanting to grab people's attention. Maybe use some sort of 3d.
00:29:36
Speaker
whatever right like ESPN the magazine I used to always kind of make fun of because it like every graph and ESPN the magazine was like a 3d monstrosity but like their goal is just fundamentally different than what at least you and I are trying to do 99% of the time right which is to make a point or to or you know it's not about getting clicks right it's about informing and educating
00:29:57
Speaker
Yeah. And the key though is really keep in mind that, you know, if you're going to do something like that, you're going to go for kind of an attention grabby design. That's fine. Just be aware of the costs, right? Like there are usually costs, you know, involved in doing that. And if the cost is that, you know, you're going to leave most readers with a just fundamentally incorrect understanding of the reality behind the chart, then maybe it's not worth it.
00:30:24
Speaker
Um, what is your thought on this, this distinction? And so I've just been thinking about this, so I'm curious just to get your take on this distinction between simplicity and clarity. Cause I, I feel people, I hear people say like, Oh, you should be able to get it in a second. Like.
00:30:39
Speaker
And that may or may not be true, but even very complex data can be represented so that it's clear, right? The data itself is not simple. If I think about
00:30:55
Speaker
Okay, try to come up with a good example here off the top of my head. A line chart that has 50 states on it, and maybe two of them are kind of highlighted, right? Like, I don't know if I would call that a simple graph, because it's got 52 lines on it, but if it's, you know, colored correctly, and you know, you've got the highlighted lines versus the gray lines, like, it's clear. It's just not, I think simple sort of undermines what's going on in the data, right?
00:31:21
Speaker
There's complexity there and it's our approach to clearly communicating it. That's the goal. Yeah. Yeah. I mean, uh,
00:31:32
Speaker
Yeah, I mean, that's a, you know, that's a great point. Like, I'm trying to remember those, I think it was Einstein quote, basically said, you know, essentially, you want to simplify, but not oversimplify or something along those lines. Yeah, by butchering it. But yeah, like, you know, because the goal, I think, I think as a general principle,
00:31:54
Speaker
we should try to simplify the charts, especially these everyday charts for reports and presentations, as much as possible. But sometimes, as much as possible, it's not very simple. Like if the the underlying message is just, you know, it's more complex, there's, you know, several moving parts to it. And if you take away some of those parts, it just doesn't make sense anymore. Or the audience just doesn't, you know, they don't get the
00:32:20
Speaker
the minimum necessary amount of information to put two and two together in their head. But I think the much, much more common problem is the opposite, where people make charts that are unnecessarily complex.

Dashboard Complexity

00:32:38
Speaker
I'm looking at it, I'm imagining a simpler version of that chart, which still accomplishes the same thing. Because kind of like what I was mentioning before, I consider that charts are graphics for doing a job.
00:32:49
Speaker
And so ultimately what really matters at the end of the day is did the chart do the job that you had in mind when you first decided to create a chart in the first place? Because we don't just create charts just to show the data, right? There's always some ulterior reason, right? We're trying to explain something to somebody. We're trying to answer a question that they've asked us. We're trying to maybe persuade them to take a course of action or adopt an opinion. And ultimately that's what matters is how well does the chart do that?
00:33:19
Speaker
And so there's a bunch of things that go into that. And one of them is simplicity, right? You want to, you know, sort of make it as simple as possible for the audience to actually, you know, understand what it is that you're trying to explain to them or, you know, agree with your opinion on whatever it is. Right. And that's hard. Like it's hard to make simple charts. Like I said, the most much more common problem is our charts are kind of needlessly complex.
00:33:47
Speaker
And that's kind of the underlying objection that I have with things like connected scatter plots and box plots, is that they, in every case that I've seen, they're needlessly complex, right? They're forcing yawnies to jump through cognitive hoops that they don't have to jump through in order to get whatever the message of the chart actually is. Right. I want to close up with just one last thing. You mentioned at the beginning
00:34:12
Speaker
that you started, your data is part of your career teaching dashboards. And this book isn't really about dashboard design. And I'm curious, I guess I'll phrase the question this way. I'll make it sort of a debatable statement. And so I'll get your take on it. That creating static graphs are harder than creating dashboards, ignoring the tools and the technical challenges.
00:34:40
Speaker
that creating an exploratory dashboard is easier than creating the chart that you just described, where you're making a point, you're making an argument.
00:34:47
Speaker
Um, I think that very much depends on what you mean by a dashboard. Yeah. You know, terms of a lot of baggage. Uh, and like, for example, in my dashboard course, I actually segment, like the very first thing I do is say, we got to segment this term. We can't talk about dashboard design best practices when dashboards, like, as far as most people are concerned, a dashboard is any display with a bunch of charts on it.
00:35:13
Speaker
Right. And so that's going to include all sorts of stuff. It's going to include basically like what you and I probably call infographics. It's going to include interactives on, you know, news sites. It's going to include like real time status monitoring dashboards and manufacturing plants and workout dashboards on your phone, like, like all sorts of different things. And so really,
00:35:36
Speaker
I have, you know, I kick off the course by essentially segmenting dashboards into nine different types. I have this kind of taxonomy of dashboards. And so in terms of like, well, is it harder to create a dashboard or a static chart? My, you know, my first sort of counter question would be what kind of dashboard are we talking about? Yeah, because there are some kinds of dashboards that are very difficult, very challenging to nail, right? Much more so I think than
00:36:02
Speaker
than most static charts. But there are other types of dashboards, like if you're basically creating kind of an infographic, like a collection of static charts on a poster or something like that, that maybe that's actually easier. Right, right. Well, that is a really good clarification. And we won't go into too much detail, because I'll just give you the motivation maybe to write your next book will be practical dashboards, maybe will be the Oh, it's in the pipe.

Closing Remarks and Contact Information

00:36:28
Speaker
Okay it's in the pipe so we've got that to look forward to so we'll get you back on the show when that book comes out. Okay so Nick last question and I'll put the links to everything we've talked about with the two Nightingale articles and I'm of course right after this uh after this call I gotta go back and read the box plot one so I can I can think about that some more. But where can folks find you how can they get in touch?
00:36:50
Speaker
The best way is just through my website, which is practicalreporting.com, all along Word. I'm easy to find on LinkedIn because I'm the only person with this name, unsurprisingly. Yeah, so on the website, you'll find information about my books and about upcoming workshops. I do have one actually.
00:37:11
Speaker
coming up starting on May 6th. About three times a year I do public online workshops. And so if people are interested in taking the practical charts course or the practical dashboards course, which is also taught during that workshop, then I would welcome them to register. Terrific. All right, Nick, thanks so much for coming on the show. The book is Practical Charts. I hope folks will check it out wherever you get books. And I'll put links to all the things we've talked about in the show notes. So thanks again for coming on the show. This was great.
00:37:39
Speaker
Yeah, really great discussion. A lot of fun. Thanks everyone for tuning into this week's episode of the show. I hope you enjoyed that conversation with Nick. I hope you will check out his website. I hope you will check out his book. And from my perspective, I hope you will rate and review this show on iTunes, Spotify, wherever you get your podcasts. It really does help me out. So until next time, this has been the Policy Biz Podcast. Thanks so much for listening.
00:38:07
Speaker
a number of people help bring you the policy of this podcast. Music is provided by the NRIs, audio editing is provided by Ken Skaggs, design and promotion is created with assistance from Sharon Satsuki-Ramirez, and each episode is transcribed by Jenny Transcription Services. If you'd like to help support the podcast, please share it and review it on iTunes, Stitcher, Spotify, YouTube, or wherever you get your podcasts.
00:38:28
Speaker
The Policy Vis podcast is ad-free and supported by listeners. If you'd like to help support the show financially, please visit our PayPal page or our Patreon page at patreon.com slash policyvis.