Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
Episode #100: Jon Schwabish
175 Plays7 years ago
Welcome to the 100th episode of the PolicyViz Podcast! 100 episodes! More than 100 guests spanning more than 33 hours to bring you the best of data visualization, open data, storytelling, and presentation skills. I hope you have enjoyed the...
The post Episode #100: Jon Schwabish appeared first on PolicyViz.
Recommended
Transcript
Introduction to the 100th Episode
00:00:16
Speaker
Welcome back to the Policy This Podcast. I'm your host, John Schwabish. For this, the 100th episode of the podcast, and as you heard, new intro music. One time only, special intro music. A friend of mine, Nayan Bhula, here in Northern Virginia,
00:00:32
Speaker
His band, the NRIs, has been around for a long time. Finally got to go see him play in DC a few weeks ago and just want to switch things up for the 100th episode. So I'll put a link to the NRIs. They're here in DC. If you're in DC, you should try to check them out. They'll be playing in Georgetown in December.
00:00:49
Speaker
So, just wanted to switch things up a little bit.
Reflecting on 100 Episodes
00:00:52
Speaker
100 episodes, that means I've had more than 99 guests talking about open data, data visualization, presentation skills, all doing incredible work in these various fields.
00:01:04
Speaker
I refer to the show task of doing a podcast as lazy man's blogging. It's a lot easier and a lot more fun to find someone who's doing incredible work, have an interesting, incredible conversation with them and pull that together rather than trying to write 800 words on a topic that
00:01:24
Speaker
you know, maybe I don't know that much about. So on this, the 100th episode, it's just me. You get to listen to me rant and rave for a little bit. I want to do a quick look back and then a look forward. Both look back on what's happened with the show over the last two and a half years or so, and then a look forward about what I've been thinking about with respect to communicating data and data visualization.
Challenges of Weekly Publishing
00:01:45
Speaker
So, where have we come from? Well, the show launched in April of 2015. My first guest was Alberto Cairo. A great guest obviously has done a lot of great work and was super excited to be able to launch the show with him. I've basically been able to keep up that weekly publishing schedule every Tuesday, save for a couple of weeks, maybe the end of the year and parts of the summer, take a break, but able to keep up that weekly schedule, which has really been incredible, both that
00:02:15
Speaker
I can do it, but also that not that there's a shortage of people to interview, but they're willing to come on the show. It's always sort of a guessing game when I email someone and I'm just sort of crossing my fingers that they're going to be willing to come on and take time out of their day to join me and talk about stuff that they're working on. So I'm really grateful that
00:02:35
Speaker
You know, I've had more than 99 people willing to come on the show and talk to me for 25, 30 minutes. And also very grateful to you, the listener, for tuning in every week to listen to these episodes. So if you have kept up over the last two and a half years, you have spent more than 33 hours listening to me and my guests talk about these different topics.
Top Five Episodes Highlights
00:02:56
Speaker
So, let me talk about the top five episodes, the most downloaded episodes over the last two and a half years. Now, I'm going to guess you're not going to be surprised by the number one episode. The most popular, most downloaded episode was the 21st episode of the podcast with Edward Tufti. E.T., as he goes, by E.T. and I met here in Virginia before one of his workshops here in Virginia.
00:03:20
Speaker
We sat down in the big ballroom with my gear and talked about all things data visualization. We talked about data art. We talked about what it means to communicate data. We talked about how he does or does not take in feedback and suggestions when it comes to his workshops.
00:03:36
Speaker
So that's number one, probably not a surprise to most of you. Number two, number two most popular show of the last two or a half years or so was the crew from the Financial Times. So I was fortunate enough to talk to Martin, John and Alan about the work they do, about how they think about getting reporters and news folks to embrace data, embrace data visualization and make it easier for them to create visualizations and how they think about
00:04:04
Speaker
creating different types of visualizations, from static visualizations to interactive visualizations. That was a great show. Third most popular, most downloaded episode was the Data and Women episode, a meetup group here in DC, also around the world, but I was talking to Brittany, Emily, and Julie about the new meetup group here in DC.
00:04:25
Speaker
And on that show we talked about the challenges women often face in the field of data science, data analysis, and data visualization. And that group is still going on here in DC. You should look them up if you're local and you should look wherever you live look for the chapter of the data plus women group. It's a really fantastic resource both for men and women of course.
00:04:47
Speaker
The fourth most popular downloaded episode was with Drew Scow and Robert Casara. This is a little bit of a surprise to me. Drew and Robert had written a couple papers on how we perceive and the process by which we view pie charts. It was two academic papers. So I was kind of surprised that this made the top five that people were so interested. Of course, I think it's because
00:05:08
Speaker
There's still this raging debate about pie charts. So maybe it's not so much a surprise, but it was a fairly research oriented discussion. So that was a little bit of surprise. And then finally, the fifth most popular podcast episode was the episode with Evan Sinar from Philadelphia. Evan
00:05:27
Speaker
Of course, there's a huge Twitter following. We talked about his process and communicating data, especially within organizations and to big audiences as well. So there's your top five. I won't tell you the bottom five, but those are the top five folks that I've talked to. But it's been an incredible experience to talk to the folks that I've been able to talk to.
00:05:46
Speaker
all the way from data journalism to people in industry to researchers to people who are doing presentation skills to people who are authors. It's really been a great experience. So let me also tell you a little bit about behind the scenes. How does the show
Technical Aspects of Podcasting
00:06:02
Speaker
work? How do I put it together?
00:06:03
Speaker
I get this question a lot. A lot of people are interested in starting their own podcasts, which I encourage everyone to do, but will also tell you it does take some work. It takes some time to pull these things together. So let me give you a little bit of the behind the scenes, how the show works. So as you can probably guess, most of my guests are not in DC. So most of the interviews are conducted on Skype. I use a little program called Call Recorder.
00:06:28
Speaker
which records the calls on video and I immediately convert them to audio.
00:06:34
Speaker
I'll often record with the video on, although you end up getting lower sound quality, but it makes for a better conversation, I think, so there's definitely a trade-off. Skype, if you don't know, does have different channels for video and for audio only, so the clearest audio you actually get is when you make a call and just audio only. Not going through video and then turning video off, but actually going through the audio only channel. But I tend to do video, I think it makes for a more natural conversation when you can see the person.
00:07:04
Speaker
So I'll reach out to people that I think are doing interesting work, that are interesting people, which is always good to talk to interesting people doing interesting work, something that I might want to learn about. As I said, it's lazy man's blogging, so it's great to just find really interesting people doing interesting stuff. So I'll reach out. I'm always grateful they are willing to come on the show. I mean, it's pretty incredible that they've had more than 99 guests come on the show and talk about the work that they're doing. So I'll reach out and we'll chat a little bit, usually via email, about
00:07:33
Speaker
what exactly we want to talk about in the 20 or 25 minutes and I'll basically try to come up with a sketch of the interview so I'll list four or five questions and of course let them chime in and we use that as really the outline for the discussion.
00:07:48
Speaker
We don't, I don't always follow it, you know, line by line because we'll have a con, we'll talk about a topic and something will take us, you know, veer off and we'll, we'll get in deep into some, some topics and maybe not get through everything or we'll get through more than we had planned. But I use that as the outline to frame the discussion and.
00:08:08
Speaker
You know, I'm not doing these interviews as a way to catch anyone or to trick anyone. So we both go into the discussion knowing where we're going to try to head. So I create these mp3s. I shift them off to Ken Skaggs, who's doing the sound editing for me. He does a fantastic job. I'm always thankful that
00:08:29
Speaker
He's able to pull out my ums and uhs and the doors closing and the ambulances going by and the clocks ringing. So he's able to go in and clean up the audio and he sends it back to me. I drop it into GarageBand. I put in the intro and outro music, maybe do some minor adjustments.
00:08:46
Speaker
throw the logo image onto the file and then put it into the blog where I'm using Blueberry is my provider for the blog itself to push it out to iTunes, Stitcher, Google Play and others and then I have to write the blog post and the blog post is really a modification in some ways of the person's bio. I try to write it from my perspective and
00:09:13
Speaker
frame it around the conversation that we are having as opposed to accomplishments the person's done. As we're doing the conversation I'm writing down lists of links that I want to make sure I link to on the show notes and I'm also writing down places where I think there are good quotes that I might be able to use in banner images or on the show notes page itself or on social media. So I'm trying to take notes of those things too and I'll go back later and go through the recording and get the exact quote.
00:09:43
Speaker
So the show notes goes up and anyone who writes a blog knows there's a lot that goes along with that. I have to make some images to go with the front page of the site and then Tuesday morning off it goes hit publish. Usually try to do a few days before but usually it's late the night before I send an email to the guest.
00:10:01
Speaker
Say the show is about to come out. Thanks for coming on. Here's a couple of tweets you can use. Here's the link. Here's some images if you want to promote the show. And most of the guests are happy to do that to help promote themselves on the show. So that's the process.
Encouraging Podcasting
00:10:16
Speaker
The interviews, of course, themselves are 20 to 25 minutes, setting them up and getting the calendar set up and having the conversation beforehand might take 30 minutes to an hour.
00:10:26
Speaker
Ken is doing the audio editing that might take him two or three hours and then pulling the rest of it together probably takes me an hour maybe maybe 90 minutes something like that so it is a lot of work I find it really rewarding I hope you find it enjoyable to listen to the show I hope you would consider doing your own podcast to highlight the data visualization work that you're doing I think more voices in the field can only be
00:10:49
Speaker
A good thing out there the podcasting world is expanding rapidly. It's a lot of fun. I'll be honest. It's a lot of fun. Okay, so that's my look back. Let me look forward. So I'm just gonna talk about a few things that I've been thinking about some things that I've been writing about and talking about with folks.
00:11:05
Speaker
And hopefully we'll get some people on the show in the next few months to help me think about some of these issues and some of these topics. So the first thing I've been thinking a lot about lately is qualitative data, how to visualize qualitative data. And I know there are a lot of people out there doing interesting work with qualitative data. Anne Emery comes to mind. I haven't had Anne on the show yet, but we'll definitely do so.
00:11:28
Speaker
You know a lot of people work with qualitative data, they're doing interviews, they don't know how or they have trouble visualizing the data because inherently it is qualitative. And to me some visualization types that people are using to visualize qualitative data, to me they're inherently quantitative. So for example a word cloud in my mind is inherently quantitative. Once you've taken that text and you've calculated the frequencies of different words,
00:11:53
Speaker
you now have quantitative data. So for me, converting qualitative to quantitative is not the core, I think, of the challenge. I think that the real challenge is you've done a series of interviews, you have all this text information, and how are you going to convey that information to an audience? And one thing that I think people who do qualitative research have an advantage over those of us who do more quantitative research
00:12:21
Speaker
is they have personal stories or personal narratives, personal experiences that the people in their studies or their research can tell the researcher about. And one thing I think that people who work with quantitative data don't do enough is to think about those stories and how we can use those stories to connect with our reader, connect with our user, make it more personal and make it more meaningful. And so I think one of the things I think perhaps qualitative data analysts
00:12:48
Speaker
Should take advantage of is the fact that they do have these more personal these more narrative pieces that they can use so pictures with images having some overview of the space that might be a little more quantitative but then to allow the user to drill in a little bit and find the true qualitative information in the series of in the form of quotes or annotation or pictures.
00:13:11
Speaker
I think is a real advantage, and it's something I've been thinking about a lot. I've been building out a collection of qualitative data forms that I've been thinking about how others have used. I think it's a real challenge, but I'm excited about the opportunities it affords us because we know how important making these personal connections is, how we as readers of content really connect, I think, with
00:13:33
Speaker
Stories with reporting that we can personally relate to that I think is different than when we see it in a form of a bar chart or pie chart or even an interactive visualization but when it's someone's experience that we can relate to I think that might bring us in in a deeper and perhaps even more meaningful way.
Challenges in Data Visualization
00:13:52
Speaker
So that's the first thing I've been thinking a lot about. The second thing I've been thinking a lot about is maps and visualizing data on a map. And I should have a blog post out shortly, and I'll warn you up front, it's a very long blog post, but that's just the way it is. I'm going along with this one instead of trying to cut it up into multiple blog posts. So we all know that people like maps, and they like seeing data on maps.
00:14:14
Speaker
I think there's a clear reason for that. People are comfortable with maps, they're intuitive, they're logical. I can see where I live, you know, on a map I can point to Virginia, I can point to California, I can point to Great Britain because we're familiar with them. But if you're listening to the show you probably know some of the issues with presenting data on a map. Some geographies are much larger than others and so they may take up a larger space on the map even though the data may not be that much more important.
00:14:41
Speaker
There are issues with population density. Montana is a large state, but it doesn't have as many people as a state like Rhode Island, which is a much smaller state but has more people per square mile, so the population density is much higher.
00:14:53
Speaker
The third issue and the issue that I've been spending a lot of time thinking about is how we choose the breaks in our color palettes when we make a choropleth map, when we use color to encode data. So a lot of the basic tools take what I call an equal classes or an equal binning approach, which is they take the range of the data, you just drop the data in, it takes the range of the data and splits it into four or five or n groups.
00:15:18
Speaker
And the data just dropped into those groups and I think a lot of people probably don't give a lot of thought to that process what's going on and the fault tools like tableau like an infogram like a Google sheets but those default approaches may not be.
00:15:34
Speaker
Presenting the data in the best way if the data are not uniformly distributed if they have a big skew if they have an outlier if there are big differences in the values they may not be representing the data in the best way so I've been thinking hard about that because there's just this inherent trade-off with with visualizing data on a map.
00:15:55
Speaker
that people like maps and yet there's this challenge with really trying to show the nuance and the detail and the differences that may not be apparent when they're presented on maps. So I've been thinking a lot about maps, I've been thinking a lot about interactive maps versus static maps as well. So something I've been thinking about and we've seen a lot of growth in that area with Mapbox and some of the work that folks in the D3 area have been doing with different projections.
00:16:18
Speaker
I'm optimistic that there'll be some more strides made in that area, but also I think that for most of us, for the person who's making a map of the unemployment rate or whatever it is every day, that there are some things that are not being thought about, that are not being considered when the map's being produced. One of the other things I've been thinking about is data visualization for social media.
00:16:41
Speaker
So we all think carefully, I think, or hope that we all think carefully. Well, let me say this. We all should think carefully about our different audiences when we're creating our visualizations.
Creating Engaging Social Media Visuals
00:16:53
Speaker
So a visualization that goes into a academic 50-page report may be much different than the graph that you put on a blog post. It may be much different than the interactive data tool that we create.
00:17:05
Speaker
But I think there's also differences in when we put graphs on social media, onto Twitter, Facebook, Instagram. As things stream by and we see quick visualizations fly by us, do we need to be thinking about those visualizations in a different way? I think maybe we do. As you probably know, I'm not a huge fan of pie charts. When I tell people it's as false, I don't make pie charts. But if I were to make pie charts, I would use three or four slices in my pie chart for the following reason. Five or more slices, really hard to see the quantities, really hard to get those quantities.
00:17:35
Speaker
And I'm thinking about a pie chart where I have five sort of equally important groups. If I had one group and I was sort of the four other groups were in the background, maybe okay. But five equal groups, I'm really not going to show that. Two groups, I'm also not going to show that because that's really just one number. So a pie chart that has two slices, it's really just one number. This is 33%. Not sure I need a pie chart for that.
00:17:58
Speaker
That being said, maybe on Twitter I would want to use a pie chart for two slices because it's very visual. We know visuals are really important on social media and maybe showing that very simple pie chart is a better approach than showing a single number. Maybe not showing the entire map or entire visualization. Maybe taking that bar chart that's oriented vertically and converting it to a column chart that's wider, which we know is also better for Twitter and for Facebook. Maybe we need to make those changes when we're thinking about social media visualizations.
00:18:27
Speaker
So these are things that I've been thinking about. I know it's things other people have been thinking about. I've been talking to lots of people about maps, about how to visualize qualitative data, how to think about new technologies and new tools. And I'm excited for the next set of episodes of the podcast.
Future Episodes and Listener Appreciation
00:18:43
Speaker
I have some great guests lined up for the rest of this year. I'll go through the end of 2017, take a little break at the end of the year and start back up in January.
00:18:51
Speaker
So I'm really excited. I'm really grateful that you have all tuned in week in and week out to listen to the show. I hope you've enjoyed it. If you have, please do send me a note. Let me know what you like about the show, how you've been able to use the information you've learned in the show. And please do rate the show on iTunes. Please do or on your favorite podcast provider. I've got t-shirts now. I've got stickers now. I actually have swag. So you can actually go to the shop on the policy of this site and check some of that stuff out.
00:19:21
Speaker
So thanks again for listening. I'm going to roll off this week with another song from the NRIs here in DC. So thanks again for listening. Thanks again for tuning into this week's episode. This has been the Policy Eviz Podcast, the 100th episode. Thanks again. We're all clamoring for the next obsession. Forging ahead even when hearts question.
00:19:51
Speaker
Regret is best spent remembering good times And progress happens When you're in between the lines Between the lines Ciao Ciao And rejoice
00:20:21
Speaker
Listen to the echo, echo of my voice Shout and rejoice Oh, listen to the echo, echo of my voice Shout and rejoice Oh, rejoice, rejoice, rejoice
00:21:16
Speaker
The finish line it always beckons And the lines of fate get duller with every second Every second Shout and rejoice
00:21:44
Speaker
Oh, listen to the echo, echo of my voice Shout and rejoice Oh, listen to the echo, echo of my voice Shout and rejoice
00:22:08
Speaker
Oh rejoice, rejoice, rejoice I've done the best I could And I've been so misunderstood I've done the best I could And I've been