S2 E2: Byte - Data Viz literacy.

0:00Hello and welcome to What So What Now What.

0:03This is episode two of season two and I'm joined by my co-host Ravi

0:07Good evening.

0:08Hello.

0:08How is how how are you doing?

0:10I'm good thanks.

0:10How about you?

0:11You're a bit west aware, aren't you?

0:13Yeah, yeah, yeah.

0:14As you can tell from my voice I've been uh struck down by the the old laryngitis, so I actually couldn't speak.

0:19for about four days.

0:20So for anyone that knows me or was um knows of um me, I that that's torture.

0:26And I I'll tell you my internal monologue has been something to um witness.

0:31Uh wish you speedy returns.

0:33I think we would try to recall this on Sunday and you you you sent me the the most uh

0:37Like I I I listened to your voicemail and I just thought, no, we're not recording this for a while.

0:43Yeah, it was pr it was proper raspberry, like this it's the the rattle of a man on his deathbed.

0:48Yeah

0:48Well, today's episode is actually about visualization literacy and it is a bite.

0:54It is indeed.

0:55And um it's quite a hot topic right now.

0:57Um so

0:58not just visualization literacy but data literacy.

1:01Um and we're gonna sort of um dig into a bit more about what that means, what what we th we we feel it means and the things we've found out when

1:08Um funny I'm um digging into this topic a bit more and of course um uh debating a bit more about our our thoughts on on the topic.

1:15And yeah, and we within that we can think about

1:18you know, what data literacy is, but what beyond that, what is literacy in general?

1:22Um, 'cause that's something that's is important, and then how does data literacy and then further beyond that viz literacy

1:30But before we do that, let's um let's first thank everyone for the feedback from last week when we talked about hyper.

1:35That was actually a bit.

1:36Absolutely.

1:37Um really good feedback on LinkedIn, Twitter.

1:40I think it was a bit dry at times, uh and you know, appreciably so, given the topic.

1:45Um

1:46But I think people really valued sort of the insight into sort of where the technology's going.

1:50Yeah, it's something that sort of um yeah, I mean we we we I think w one one listener noted that we said it would be fairly accessible, but

1:58E I I won't lie, uh even at times with that that final section, there was bits where I was just completely like lost and I was like I

2:05Yeah, I mean this this makes sense in a weird way, but I couldn't like explain it back to someone else.

2:11Um but that's absolutely fine.

2:12I think we're gonna

2:14Um do uh try to tone that sort of l um heaviness down in this in the in this episode for sure.

2:20Absolutely

2:21And the other thing we we got good feedback on is the um show notes and also the um the podcast uh artwork.

2:29So if you if you've noticed in the previous episode and possibly in past episodes

2:33As we go through various sections of the show, you might notice that the uh podcast album art changes.

2:40And usually what we do here is we show images relevant to what we're talking about.

2:43So we just wanted to call that out.

2:45It's um it's a feature that's supported in Apple Podcasts.

2:48and most other podcast apps.

2:50So just um bear in mind if you're listening with your phone in your pocket and you're wondering what we're referring to, um it's probably the image uh of the album art.

2:59So just switch on your phone and have a look

3:01And and if you're someone like me, for example, who like listens to on the train and why I'm with my eyes shut or uh you know while doing other things, that's absolutely fine.

3:09I mean you can also scrub, so like hold down the tracker and

3:12um see what what changes and then listen to the clip to get a gist or recall on what we are talking about in that in that point.

3:19I really like those flags.

3:20Again, hats off to Tim for all the hard work he does uh on the post-processing.

3:26I think we maybe we should uh talk a bit more about that at the end of the episode.

3:29Yeah, yeah, why not?

3:29I think at the end of the episode we'll touch a little bit on how we actually record the podcast, how it comes together, what tools we use, and how we host it.

3:36Why not?

3:37Yeah.

3:37Yeah, cool.

3:39Cool.

3:39So visualization literacy, what what does it mean?

3:44What do you understand by that term?

3:46I I think it really comes down to how people

3:48when you sort of look at a chart or a graph and even, you know, as as basic as do you call it a chart or a graph?

3:53When is a chart a chart or when it was a graph a graph?

3:56Stuff like that, it's like

3:57How does that terminology and all those different things fit together?

4:00Um and if you think about literacy in general, you know, the ability to read and write

4:05um that's taught to you or you pick you you don't really pick up the ability to read right it's it's a almost like a muscle reflex that you have to teach yourself.

4:14You know there's um

4:15Always stories about those folks who you you grew up and they they just never learned to read um or write because they never needed to, but they're you know super hard on maths for example 'cause

4:25while them they never recovered the muscle reflex to write, um the day they they their mind is fairly well developed.

4:32But

4:33That that sort of holds true in when you think about literacy broken down into the in terms of data and viz, right?

4:39So if you took in thinking about data

4:41I put out a tweet recently which was um is data mathematics?

4:45Uh yeah.

4:46I'd argue not, right?

4:47Um but it it's all about those uh what you associate with that and how how you sort of

4:52approach to it.

4:52And so I guess similar to visualization, like is an infographic of his, you know, what is what is it that you see when you look at an infographic or something very complex versus a bar chart or a line chart

5:02Yeah, and it's it's i it's almost taking literacy in in its in its simplest form.

5:07So uh traditional literacy is is about sort of the ability to read and write

5:13Uh data literacy is is some of the same, but you're doing that with data, however that might come.

5:17It might be something as, you know, hard as numbers in an Excel sheet or

5:22something as beautiful as like a a hand-drawn data visualization, you know?

5:27Hard here meaning static, the difficult.

5:30Exactly, exactly.

5:30And then then there's also the other sort of

5:33deeper sort of form of literacy which is not just the ability to read and write but also then comprehend and understand.

5:40So you you you talk about comprehension at school.

5:44And so data viz literacy is kind of an amalgamation of these core skills sort of coming together to form something quite abstract actually.

5:53It's a very difficult thing to sort of appreciate and and score.

5:58Um but actually a few people have actually tried to do this.

6:01Yeah, exactly.

6:02And it it's funny when we think about this as uh a hot topic right now.

6:06Um

6:07But really when you one when when you know are you you start looking at data literacy, vis literacy given the fact it's a hot topic and you start looking deeper into it, you figure out actually this has been going on for

6:18At least five, maybe six or seven years.

6:21Uh and you know, some of some papers are even older than that, where people have actually sat down and thought about how do we perceive these different elements that we interpret

6:30Now of course when many in who work as intensely in our fields as we do may know the names of Stephen Fugh, Alberto Cuaro, uh David McCandless, Edward Tufte.

6:41um those sorts of people and th they they sort of touch on all of these elements when they're telling and sharing their their particular frameworks as it were.

6:51Never a right and wrong way of course on then all these things.

6:53It's all about the application.

6:56But all of all of what they've shown has roots in this what what really is a almost a gateway skill, right?

7:02Like you you think about um data literacy and vis literacy um

7:07But more in in in particular viz literacy you you're thinking about n numerical skills in in some ways.

7:12You think about the computer skills, the ability to design something.

7:16And even he and you know, touching on design, you've got stuff like the the media literacy

7:21And you've got uh you know the ability to design something and build something that's engaging as well as informative.

7:27Exactly, exactly.

7:28And so it's a really difficult thing to score because I think most people are probably listening to this and saying, okay, fine

7:34Uh we understand the abstract sort of nature of it, but how on earth do you then start to quantify it in a way that, you know, is is meaningful so that you can actually understand whether people in your organization or people you're engaging with

7:47um sort of what their level of literacy is.

7:50Um and there's a couple of really, really good papers on this.

7:53I've I've highlighted one in the show notes

7:56Which basically breaks it down into, I guess, levels.

8:00And the the most basic level is the simple sort of extraction of information from data.

8:05So that's basically being able to look at numbers on a page

8:08and you know take that data in.

8:10Then there's the intermediate level which is about sort of seeing trends and understanding relationships.

8:16So for example, correlation when this goes up, this goes down.

8:19That's sort of an intermediate level.

8:21And then the advanced level, the sort of the really deep level of understanding, it comes through comparisons and and comprehension.

8:29Basically being able to look at data

8:31understand the trends and then relate some sort of higher level of meaning or sense from that.

8:36So uh for example if this increases then the consequence of that might be X.

8:41And that's that's the kind of

8:43really advanced level of comprehension.

8:45And so it boils down to a really basic set of things, you know.

8:49Can can the user take some information out of out of the data viz as a start?

8:54And is it correct when they do that

8:57Um when they look at that chart, um uh are are their sort of thinking process activated.

9:03Sometimes you can look at something, but it doesn't get your sort of brain juices fizzing, as it were, which is which is actually a really important point we'll come to later.

9:11Um, you know, the things that I've been emphasized in the graph, whether they're outliers or, you know, things that don't fit the trend, are those being perceived correctly?

9:20And then once you've done all this, are you able to to sort of chunk it up into sort of visual pieces and then build a story?

9:27That's that's a very big word that's often used in this space.

9:30Are you able to then start to build a story?

9:32And that's why storytelling through Dataviz is a really important thing because if you can help the user do that, then you're you're reaching this sort of really high level of understanding.

9:41And then the last thing is

9:43I I'm I'm surprised this has lost, but it's actually really important.

9:46As the user walks away from this, you know, let's say in a week's time on two weeks' time

9:51Are they still uh able to remember what it is they digested from your visualization or from your your piece of data, right?

9:58Say uh does it have a lasting impact, right?

10:01And actually, I don't know, Ravi, when was the last time you looked at a database?

10:04Uh yeah, probably uh earlier on today.

10:06Okay.

10:07And do you remember the numbers that you took from it?

10:11Um I I I honestly cannot.

10:13Okay, but what do you remember about?

10:15Yeah, so it was uh it was a visualization about uh Peter Crouch and his scoring for his last three clubs in the Premier League.

10:22Okay.

10:23Cool.

10:23So that's that's it.

10:24I c I can't tell you whether it was what what the axes were, what the trend was, but I can just tell you that those are the key things that were in that.

10:31And what did you what did you take away from it?

10:33Um not it was sort of like yes, it was almost like a okay, and then you just move on.

10:38There was no point, yeah?

10:39Yeah, it was sort of uh almost a supplement to the wider story.

10:42Okay, and so that's that's really interesting.

10:44We'll come on to this example at the end.

10:46But the thing you took away from that visualization was just the fact that it was a visualization about Peter Crouch.

10:52That was it.

10:53There was no salient point that stuck with you.

10:56And so that's a very, very good example of actually what goes on with every single data visualization.

11:03But you don't actually conceive that whilst you're building it.

11:05I bet you when you're building something or I'm building something

11:09We build this sort of I'm gonna say fanciful story in our mind about what people are going to understand and how much they're gonna remember it after

11:16But in real terms I ask you what you you saw earlier on in the day and you can't even tell me, you know, a a pertinent number from that thing.

11:24Yeah.

11:24That's so that's so true.

11:26That's absolutely true.

11:27And and I th I think this is something I learned so uh like not not so early on, but like the blood, sweat and tears you put into a piece of work in the back end, the the middle end, and then when you deliver it

11:37And you you you try and tell someone like, oh, so what I had to do here was build this really complicated calculation.

11:43Uh the the data back in is a mess, so I had to clean it up in this really fancy way.

11:47I think it was a really inventive thing I had to do

11:49But at the end of it it's just gonna be an image on a slide deck.

11:53Yeah.

11:53A slide deck which is eighty-three slides long, um and it's gonna be flicked through after the one salient point has been made.

12:00Yeah.

12:00And you're just thinking, wow, I spent so much time bas uh purely because of

12:04you know, the trust and the rigor that that uh applies behind, you know, like the fact that it has to be right should anyone question it.

12:11Yeah.

12:11Um but that that's that that's so true.

12:13Like

12:13Uh I mean th the the the calling me out there is I I feel kind of bad now for the person not creating it because it's now that was completely unscripted folks.

12:21Like I had to really think about what I'd looked at.

12:24And yeah.

12:25Um

12:26I mean for for me and w when you when uh as as you said to him, when you're designing this stuff you're thinking, oh so I'm doing this specific design choice in order to

12:35nudge the user to think about this and and at this level of interactivity when really if we if we're completely honest if we create interactive visualization I'd even hesitate not even five percent of users uh will click through.

12:46Yeah, exactly.

12:47Exactly

12:47That that's the that's just fundamentally what it is.

12:49Unless it's something like that has to be.

12:52Like you have to click through in order to get any sort of information.

12:55If you can get it from the static image, you sort of consume and move on.

12:58And and the thing the thing is is like you were looking at a fairly, I'm gonna say, um, extracurricular database.

13:05You were kind of just interested in sport, you're looking at this.

13:07So there isn't as much pressure, for example, as you would have in in say a business context where

13:12you know, people need to understand the numbers and they need to get the salient point from these data visualizations.

13:18And so how do you make sure that comes out in a reliable way, not just to one person, but to everyone in your organization who has access to something like that?

13:27And it it it's exactly that and like how much time are they gonna spend and that's where the lit uh the visualization literacy point comes in, right?

13:34You have to really understand

13:36At what level are all your users at and at what level are people going to interact at it?

13:41Exactly.

13:41Yeah, exactly.

13:42And and this is where things get really interesting.

13:45Um

13:46I want to sort of pivot now to the so what.

13:48So now we've we've we've sort of broken down this literacy, we've broken down why it matters.

13:52Um so the so what point here really comes to this point we've just highlighted, which is

13:59As designers, as we build data visualizations, as uh technicians, as I might say, or developers as we're sometimes called.

14:05I hate that phrase, developer.

14:07I don't code anything yet.

14:09No.

14:11Um but as developers or technicians, whatever you want to be called, um we

14:16We sort of get carried away in our own skill set.

14:20And I I think I think this is this is more so, especially when the tools are good.

14:24When the tools are good and easy, it's easy to build more complicated charts or more complicated designs or

14:30No, really, I'm gonna say extravagant visualizations that we think do a better job of showing some sort of um behavior.

14:39And so there's a study that was done in the US um where some researchers essentially took

14:44um a set of visualizations to museum.

14:48Okay, and the museum, a science museum is a fairly self-selected crowd of people who you would think have a natural um interest in sort of

14:56information and data, right?

14:58Um it's not the the the these are these are sort of people who will who will go on a Saturday afternoon to spend the entire afternoon wandering a museum.

15:06They're not gonna go there for about twenty minutes and pop up again.

15:09Yeah, exactly.

15:10And

15:11And so what they were doing is you know they they put a set of visualizations in this very simple grid.

15:16If you're looking at the uh um podcast artwork, you can have a look at the image visualizations, and it's a real range

15:22Some of it as simple as a map, the classic chloropleth map.

15:28You've got um small multiples images, you've got a few networks diagrams and Sankey diagrams

15:34A few tree diagrams, uh some standard bar charts, you know, with the 50-50 split that you often get with jet uh on gender.

15:41Yeah, the there's a underground, London underground map.

15:44Yeah, exactly, exactly.

15:45And there's a word cloud, okay

15:46And a simple line chart that was probably 100% done in Excel, but we'll we'll skip past that.

15:52Um and so this is a really good

15:55range of visualizations.

15:56Actually these visualizations were sort of headline bits of um you know data storytelling in various from various sources around the world.

16:05Okay

16:06And um what the researchers were trying to do is basically show these to people and number one understand whether they they knew what they were looking at, could they identify the charts.

16:15Um could they tell the researchers more about sort of where they were most likely to interpret data viz?

16:22And I think um what was also interesting about this is that there was a very even split between young and old participants.

16:29So

16:29It's a r nice good cross section of um of of ages and demographics as well.

16:35And um what was really interesting was sort of the outcomes.

16:39And so I think the first most pertinent sort of outcome was this

16:43uh an understanding from the researchers that people were mostly experiencing datavers or charts or graphs in a school setting.

16:52So the most

16:53prominent exposure to data visualization was actually at school or high school.

16:59And then that that was that was about fifty percent for youth, and for everyone else it was mostly in the news and so on and so forth when it came to that youthful age.

17:07Yeah.

17:07But that sort of makes sense, right?

17:09Like i y you you you and I like when we're in school we we learn about the I I I I always say like the first chart line is the bar chart, then the pie chart, then the line chart.

17:18And then maybe you do the scatterplot when you're in like high school looking at like

17:22Y y equals x squared and all of that algebra stuff, right?

17:25Exactly.

17:26Um but but I mean then then beyond that if you don't choose to work in uh any sort of analytical

17:33career then yeah you you're gonna and your your primary source of um inter in interpreting data and data visualization will be the few the media and and other such outlets right

17:45And what is then interesting is that the adult uh perspective this is very different.

17:49I think it's um you know, schooling is still

17:53uh a a a serious chunk about thirty percent but now other sources sort of start to be more prominent.

17:58So twelve percent news and media, another twelve percent at work.

18:02And then um other contexts might include books, so uh dataviz is often used in books uh as a way of communicating information.

18:11And so, um, you know, about 60% of people are experiencing dataviz either at school, uh, at work, or in other literature.

18:21And through this exposure, they did a really nice sort of uh matrix of people's understanding of different chart types.

18:29And they basically asked people if they could correctly identify chart types, right?

18:34And it it's it's it's fascinating because if you take a simple chart, uh a simple line chart

18:42Only sixty-three percent of people could actually identify it correctly as a line chart.

18:48Okay.

18:48Um you you even had numbers such as ten percent thinking it was a map.

18:54Which is which is which is really interesting, okay.

18:58And then if you go up to a graph like a typical bar chart

19:02That has quite a high level of understanding, but even that, only 75% of people correctly identified it as that, as a graph.

19:10And it's absolutely fascinating because we we often think s something like a chart or a bar chart are the simplest forms of charts you could possibly use.

19:19Like how can someone not understand a bar chart?

19:22But apparently that's the case.

19:23But I think there's there's there's other elements within this, right?

19:26So it's it's not just the fact that um it's

19:30It comes down to if you don't have to use that muscle so regularly.

19:34Like you know, similar to what we're talking about with reading and writing.

19:36Yeah.

19:37Yeah.

19:37Um think about write writing in different languages.

19:39So um

19:40Uh I for example uh speak and write in a second language which is Gujarati, uh which is the language uh that my parents use we speak at home.

19:48Um

19:49If I don't regularly write in that language, there can be times where I sort of forget the right sort of um it's not really a letter, it's like the um the the uh symbol we use uh and in the correct usage of it

20:02Because you have to really think about it and again I guess it this almost follows the same sort of mental process where you have to think really think about what am I looking at and

20:12you know, we we we we looked up here and sort of said that line chart's definitely an Excel and I think that's also something to not not really turn our nose at too much either.

20:21Because we I mean me me and you both both started in Excel.

20:24Um that that's actually where most of

20:26I'd I'd say almost everyone has created a chart.

20:29Yeah.

20:30Um if they've created a chart in a computer program, that's it.

20:33That's where they would they would have done it.

20:35They wouldn't have chosen Tableau, for example, they wouldn't have chosen Power BI.

20:38as a default.

20:39They wouldn't have gone and learned R to um use ggplot, right?

20:43To to create a line chart or a bar chart.

20:46So it's it's it's really interesting to always see people be like, oh we're using it, so you should really get yourself

20:52onto one of these more beautiful more beautiful looking tools.

20:56Going back to your point on, you know, this the naming thing is so important.

21:01Yes.

21:01Like how how do we name things and how do we um

21:05sort of approach that I think and the the w further in the study we look at um uh the network layout um chart.

21:12Yeah.

21:12Or and it's like

21:15How many people called it a network layout?

21:17Yeah.

21:18And the the the answer was uh 2.

21:206 and the most popular answer of course was it's a graph.

21:23Yes, like that's that that's what it is.

21:25Yeah.

21:25Um I I can't tell you what type of graph it is off the top of the

21:29Oh off the top of my head, but it's it's it's a chart.

21:32Yeah.

21:32And i if you look at the whole spread of all participants, um uh the the total sort of uh percentage that got that right was zero point five percent.

21:41So it's uh

21:42It's uh it's simultaneously the least recognized and always the h and also the hardest one to understand, if that makes sense.

21:50So

21:50It's a double whammy.

21:52Not only do people not know what it is, but it's also the hardest one to read and understand.

21:57Yeah, and this the we'll we'll get into more about into this in in the in the now what sort of section where we talk about, you know

22:04Um not not quite viz crimes, but things that you have to really consider when you're creating something.

22:09You know, the thing we talked about where, you know, when I look at a chart, what am I going to actually retain from that?

22:14Um and then also uh we'll we'll touch on like the ways you can make sure someone retains a certain part of information, right?

22:19Exactly.

22:20Um because with a network chart, as you say, it's really hard to interpret, but

22:24The reason people like it is because the thing it does is it shows relationship.

22:30Yeah.

22:31And in in in and it adds a spatial element to it.

22:34It shows transfer and movement, doesn't it?

22:36Yeah.

22:36Yeah, and but if you if you immediately don't understand that, you don't get it.

22:41But this is another one of these things where it's like but when you take the time to understand that

22:45And you stand there for about, you know, a a couple of minutes really digestive.

22:49And then you start exploring it visually, then it starts to make more sense.

22:53But again

22:54That that this literacy is not only just about understanding things, but also having the patience to allow yourself to understand it.

23:00And it's also about whoever's building it, bridging that gap.

23:03So if you are going to do something like a network layout

23:06Well, expect to maybe educate the user on what a network layout is.

23:12So maybe give them a legend or some guides or

23:14or something like do like a small network layout diagram just to show them how it works with something very simple and then introduce them to the more complicated one so that it doesn't hit them so hard so not only are they having to understand the topic but also something else.

23:29Uh but we'll come on to this a little bit later on actually.

23:31Um we've got some great content we're gonna signpost people to about what actually matters when you're building a viz.

23:38But just to sort of close off this section, I think the other thing about this is that um when people are looking at visualizations, there is a social element that's not often talked about

23:51If you to do a very simple thing, which is you take a uh a scatter plot, let's say it's an another sort of contentious chart type, and you show it to just one person.

24:01Um that one person in isolation might find it harder to assimilate and understand that scatter plot compared to if you show it to a room of maybe 15 people.

24:12And the reason here is that um there's a ri there's a really famous um project run by the New York Times where basically they take uh data viz from their news stories and they educate children through it.

24:23I'll put a link uh in the show notes to it

24:26And what is really interesting and they started noticing is that where students didn't have an understanding of a topic, they would look at their peers and see what they were saying and how they were reading it.

24:37And then in subsequent exercises you'd find those same students borrowing language from their peers, uh if that makes sense.

24:44So it's this idea of

24:45You know, if you have champions in your organizations who get this, actually immersing people around them or putting those people into situations where maybe others who aren't as familiar with dativiz

24:56And that kind of knowledge could rub off because people are learning from each other.

25:00That sort of social learning um sort of happens in a in a in a in a collaborative way

25:06And that actually lifts everyone's ability to understand that.

25:09And there's nothing wrong with with that, of course, yeah.

25:12Um it's just a another type of understanding.

25:15Exactly.

25:16And and and and uh you know, w I I guess me and you both see that first time with with you know the four months at the data school because it's like when I personally came in and w I think we talked about in the previous episode about

25:26And you know, learning and teet being taught tableau.

25:28Discovery.

25:29Yeah.

25:29Discovery tab yeah, discovering tableau.

25:32And for me it was like regular hammering of the points of st things such as, you know

25:36uh Andy being like, you know, remember about the context you want to show, like framing this in a certain way.

25:42Remember to, you know, annotate these things correctly in order to shape the story and use these elements within the chart.

25:49Really

25:50you know, in in a different way just to help your user.

25:53Like that sort of thing you don't learn until like you're not no I don't want to say corrected.

25:59I want to say nurtured.

26:00Because it's it's one of these things about feedback.

26:02It's like you don't

26:03give feedback without request.

26:05You say you sort of almost say, would you like some feedback?

26:08I think um Fee Gordon did a really good um piece on this in her talk at T C.

26:12It's like you do the

26:14good feedback, um here's something you can improve on.

26:16Yeah.

26:17Then but overall like it and then that's the sort of that repetit repetition builds that muscle back up.

26:22Exactly.

26:22To help understand

26:24Um where someone's coming from.

26:25And actually also when you see a community of people doing the same thing, you're more likely to take an interest because there's sort of two sides to it.

26:32number one, you don't want to be left behind and number two, you want to be part of it.

26:36I think that's, you know, the one of the strongest aspects of um initiatives such as Make Over Monday or or Viz for Social Good or any of these sort of

26:44um, you know, data kind as well, if you've if ever come across that in the UK.

26:48Those are all using the same sort of peer collaborative

26:52involvement sort of and gaming it in a way as well to make it interesting and people are engaging with it um in that way.

26:59And what I really like about those is it's not just the fact you're in a community where you're learning and

27:04almost learning and copying of each other and like developing that.

27:07They're also being challenged, right?

27:08They this is what's moving this field forward.

27:10This is why we always like come back to similar arguments about pie charts and whether a truncated axis is a good or bad

27:17And you know, this is why we continue to have these debates to make sure that we're re-educating and relearning these muscles to think and challenge the way we do, just so you're not flat out shutting someone down.

27:29Exactly.

27:29Exactly.

27:30Okay, so I think I think we've we've we've ringed that top part of the topic out pretty pretty well.

27:36So what we now want to do, now we've talked about this topic in in depth.

27:40We want to sort of enable you to start building your visualizations with certain factors in mind.

27:45And so we're we're going to lean on some work done back in 2015 actually by Helen Kennedy and Andy Kirk on a project called Seeing Data

27:53And this is a very interesting question.

27:57Yeah, yeah.

27:59And basically what they did is that they they had a slightly different approach to some of the examples we've talked about.

28:05But they run a series of focus groups and they basically uh show people some visualizations and ask them some pertinent questions.

28:13I'll actually link to a good uh podcast episode on that particular study.

28:17You can listen to it on uh data stories.

28:20But um the the most interesting part of this is actually the takeouts from it.

28:25So they they come up with like a three part

28:28blog post which basically talks about the factors that most influence sort of what people actually remember in a visualization

28:38And um I'll sort of st I'll sort of start this backwards by actually going to the conclusion first, which is the key factors that um, you know

28:46uh sort of move us when we look at a data visualization.

28:50So I'll just list this out and then we can go through them in a bit more detail.

28:53The first one the first one um that ranked the most important is subject matter.

28:58The second was trust

29:00The third was time.

29:02This is time to sort of understand what you're looking at and read it and understand it.

29:06The second from last was confidence and skill.

29:09So people's sort of innate ability to look at the data and take it in.

29:14Uh and then emotion is the very last one.

29:17And this this is partly about design, but it's also just generally how a dataviz sort of or visualization makes you feel.

29:24So uh I don't know, Ravi, what do you think about those

29:27I mean I I love this stuff, right?

29:28Like it's uh um it I I I given I study economics and sort of had a focus of on behavioral economics in my final year, this sort of stuff is really

29:36Interesting to me because it it it focuses on stuff like nudges, which I'm as you know, I'm a big fan of.

29:42Um so let's start off with subject matter, right?

29:45So Yeah

29:46Um, as you said, for example, let's go back to Peter Crouch again.

29:50Um, it's something I'm interested in, uh, but as a less casual observer.

29:55So

29:56I get what the premise is, I get what they're trying to say, I move on.

29:59I've got the information I need, I've moved on.

30:02Now, if it was a subject matter like, say for example

30:05I don't know, uh the prevalence of horoscopes and the how successful it was in predictive weather.

30:10I would probably not even give it that much interest

30:13Um so you know you you you you try less when you're uh less invested in finding out more, if that makes sense.

30:18Um if you have a vested interest in figuring out what the subject matter has and where it places

30:27within your reader um or even yourself.

30:32It it changes the way you interpret of this but also design of this because

30:36That that a layer of context and that layer of information about the type of person you're designing for and the type of person that will consume it, that really matters when you're designing this sort of thing.

30:44Exactly, exactly.

30:46And I think

30:47The the most interesting the the most interesting thing here is the thing you took away from that thing you saw earlier on today was the fact that it was about Peter Crouch, and that's still with you today.

30:58And that's actually really important because that allows you to go back to it, right?

31:01You know that if you need to understand that topic again, just remembering the subject matter and where you got it from is enough.

31:09because you've sort of signposted and and and flagged that particular bit of content.

31:13Um I think the next thing you touched on there was trust actually.

31:17So

31:17You you had a look at it.

31:20Yeah, exactly.

31:22Uh what was the source actually?

31:24Who who produced the Viz?

31:26Uh so it's Sky Sports, I assume the source was um from Opta.

31:30So Okay, great.

31:31And so Opta is a well-trusted source, right?

31:33So we you wouldn't invaste your time if that came from

31:37uh a source that wasn't optified.

31:40But but then but then at the same time I assumed it was, given that it was on a media channel and

31:45I suppose it's probably from there.

31:47So right.

31:48And and so that that matters as well.

31:50And I actually thinking back, like when you look at a news article, if there's a stat that I'm

31:56I feel is a bit sheepy or you know a bit woolly and I'm not certain how accurate it is.

32:01I definitely look at what the source is.

32:03That's when I would that's when I interrogate the source more than

32:06if it's something I just sort of either it follows the narrative of the story I'm reading or the something someone's telling me.

32:13If it's

32:14something outrageous or something that sort of shocks you or makes you think a bit more, that's when I'd more intro I'd be more interested in the source.

32:20I think that's correct for almost everyone.

32:22Yeah, exactly.

32:23Exactly.

32:24And and then actually there's also other things into that.

32:27The reason the reason you're more likely to look at it is because uh Sky Sports or Opta haven't abused your trust in the past.

32:34So you you went in with an assumed level of trust because in the past

32:38They haven't abused that trust.

32:40They haven't done some disgusting 3D pie chart or um I say disgusting for some people that's appealing, but that's a that's another topic for another time.

32:49Um but you also trust their quality and the fact that they're very honest about what they're trying to represent.

32:55They're not trying to mislead you.

32:56And so uh whether it's

32:59trust conveyed by what you see, or trust assumed because in the past, you know, you have a sort of a good history of doing things that way.

33:08Then that's a really sort of nice um

33:10sort of understanding.

33:12Um But I also think there's also VizTrust, right?

33:15If if someone shows you um what you and I would call a really basic dashboard, you know, the sort of um the the dashboard that has

33:23four elements, three elements, you know, a map, a line and a bar.

33:27Um showing you one you know, it it's just telling its own story.

33:30It's not glammed up, you know, nothing's nothing's been cleaned or designed to be perfect.

33:38Your immediate reaction is to like I don't know.

33:41I I I don't know if I I want to trust it or not, but but but but when when you see those sorts of dashboards presented

33:48And the story just makes everything that's r related to design just disappear because it's so interesting.

33:56That's w I think that's what really

33:59That's where it's really interesting to see how people's trust develops as they sort of learn more and they figure out what their their interest is are towards.

34:07Like for example, my

34:09I wouldn't say disdain, but my lack of interest in long form visualizations is, you know, I I've talked about it a f a fair few times.

34:17Yeah.

34:17Um

34:19But that's my personal choice whereas some people really like that sort of sc scrolly telling format.

34:25Um that's fine.

34:26Yeah.

34:27But I feel like that that also f feeds into this trust point.

34:30Like what do you trust and what is it that triggers that teeter?

34:34Exactly.

34:35And I I think

34:36It's so important that people don't abuse that.

34:39Um whatever they're doing.

34:41Especially if you have a a rapport and you've got momentum.

34:44It means you have to spend even more time making sure that you build that trust and you invest a little bit of time uh creating or cultivating it

34:53Yeah.

34:54The other thing is time.

34:55Where did you see this post?

34:56Was it on Twitter?

34:58Of course.

34:58Of course.

34:59Okay, so when you look at a viz on Twitter, how much time do you realistically give it to sort of read and understand it?

35:06So typically uh I'll I'll sort of it it will be uh if it's one way it will be a tap in and you sort of look at it for about I'd say maximum ten, fifteen seconds.

35:15If it's interesting

35:16I'll tap into in Zoom.

35:17Yeah.

35:18Right to actually figure out what's going on.

35:19Yeah.

35:20Um I a really good example of this is uh actually a long form viz, uh by Ben Davis, where he looked at um

35:27uh Premier League scoring rates and comparing Harry Kane to Alan Shearer.

35:31And that was just compelling because at the face of it, I mean again, because it's scrolling thing you can't actually see all the information in detail.

35:38You can just see the design.

35:40Yeah.

35:40But

35:41the big titles compelled me to look into it.

35:44And again, I I feel like the time element is important.

35:50purely because it's it dips so much depends on how much attention you're grabbing or whether that person actually wants to look.

35:58And that second point about whether they actually want to or whether they care to

36:02uh makes a difference because it's not just about the visualization design, right?

36:06It's also about the elements within this.

36:07So things like, you know, the sorting of dimensions.

36:11Like if you're looking at

36:13Um trend over time.

36:15So you've got sales or I don't know number of um music records sold um from 2010 to 2019.

36:23But it's like showing upward trend, but it's going two thousand and ten, two thousand and thirteen, two thousand sixteen, in a in a ju disjointed order.

36:32But because you look at the first

36:35First uh element, you got last element, you've got twenty ten, you got twenty nineteen.

36:39Your mind, because this is what our minds do, fills in the gaps.

36:43We don't actually interrogate that further because like I said

36:46It's all about this thing where we will invest as much time into something depending on what our vested interest is and whether we're trained to look in the right places.

36:56Or I I don't I don't want to say trained either, whether we

36:59We would think to.

37:00I think that's that's the more salient point here, whether we actually think to um double check these things.

37:06Yeah, exactly.

37:07Exactly.

37:09So the next one is confidence and skill.

37:12I said that this is different, right?

37:13So in your example, I think you're you're quite a fay with um

37:18sort of sports data vid so you're you'll naturally have a lower sort of um well you you have a lower barrier to entry as it were to to sort of this kind of content because you you're kind of part of the natural discourse that takes place.

37:32Whereas if I took that same viz and I showed it to someone who maybe didn't understand football so well, there's so many levels of understanding there that's implied through the chart that's not necessarily uh a you know

37:45Dile directly explained.

37:47Imagine having to explain the game of football before you showed someone this chart.

37:51You wouldn't even explain the subject

37:58So the subject is is is the works both ways, right?

38:00Because you might be expert on the topic but not understand the chart.

38:03Yes, exactly, exactly.

38:04And so confidence and skill comes into this because if you're confident

38:08then you'll actually work through it and you'll figure it out.

38:10But for those who aren't, you need a you need some sort of soft introduction.

38:14And then likewise your skill level

38:16If you naturally have a high level of skill then you might find it easier even if you don't have the confidence to sort of comprehend this and say

38:24Those two things are kind of uh it it's it's a difficult one to balance up, but in organizations we we take for granted actually that not everyone actually has the skill set to understand complex uh chart types.

38:36And so

38:37Uh like we suggested earlier, having that community makes it a little bit easier and brings those barriers to entries down, but also allows people to build their skills in a supportive way where they don't necessarily have it.

38:49And and the medium really matters as well, right?

38:51So if someone expla if someone built uh showed me a network chart while they're explaining a story about um you know the the fact that they took these all these different parameters, put them together and this is what they found, it's like

39:03Wow, that's amazing.

39:04But you take that completely out of context and it's just like, right, sorry, what am I looking at here?

39:09Like where do I start?

39:10Exactly.

39:13And and that always happens with like

39:15Charts with colour explosions on them or charts where there's loads of different layers um going on.

39:22There's loads of text versus the amount of charts and those sorts of images versus the

39:26Amount of information.

39:28Yeah.

39:28And that that that leads nicely to design actually, which which is mostly encapsulated by emotion, but

39:34I mean, it's funny that design hasn't been uh a pertinent point in all of this, right?

39:40Nowhere nowhere in all this research has

39:44great design been an outcome of, you know, great understanding of data viz.

39:50But I think when we talk about emotion

39:52That is where design plays the biggest part.

39:54You talked about Ben Davis's um sort of long form dataviz.

39:59And you see, that is an example where design, and not in the literal sense of you know, colours and graphics, but

40:05by choosing to make it long form, he naturally sort of lowered that skill sort of confidence gap because he made you

40:15uh assimilate it at a certain pace but making you scroll down is making you taking information in at a certain pace which means that you're taking it in and getting an education

40:24And that is that is one advantage of the long form, is that you can actually break a story down and as someone engages with the content at their own pace, they're understanding it.

40:34And you know, two people will scroll at two different paces.

40:38Yeah, and I I I like that.

40:39I mean I I do I do I do completely appreciate that.

40:41Now what I also liked about that particular viz as well is before I double tapped to actually understand it

40:48I could see the salient points, I could see the titles, I could see the the brick lines.

40:52Yeah.

40:52I could see the images it used.

40:54You know, Alan Shearer's famous ha one hand in the air celebration, right?

40:57It's that sort of

40:58That's the emotion, right?

40:59That's that's exactly that.

41:00Exactly.

41:01It's like when you're using the right colors, it's when you're using um subheaders and you're using again coming back to this concept of nudges, the small, subtle design tweaks.

41:09And

41:10And at this point, it's like the perfect time for me to give a huge shout out to my absolute favorite blog series of all time, which is Andy Kirk's The Little of Data Viz.

41:21Yes.

41:22Um where he calls out the tiny elements in people's visualizations that he really likes and show shows why they're important, you know, like adding a parameter or a filter within a body of text.

41:33Yes.

41:34Or um

41:35the most simple is, you know, using colour and titles to refer to chart uh elements.

41:40We'll put that in the show.

41:42I love that.

41:42Yeah.

41:43Yeah, absolutely.

41:43And it's it's so important because

41:45Although it's the last thing we're talking about, uh it's the first thing people notice when it's done wrong.

41:51So when when type and font is inappropriate or hard to read

41:56when colour is sort of garish or wrong or misleads people.

42:00That's what your eyes are naturally geared to notice first.

42:03So when you don't do it properly

42:05People spend more time trying to decipher what's going on than than than they should.

42:09But when it's done so right, that stuff just sort of fades into the background.

42:14And you don't talk about the font, the colour.

42:16In fact, more often than not, you look at it and you realize, what is it about this?

42:20And you can't really put a finger on it.

42:22But that's because the design is doing the intended job.

42:25Correct.

42:26And I think when while Owen saw Andy Kirk talk and he uh put that famous hurricane visit, I think you've seen the one it's animated

42:33Yeah.

42:33And it shows the swells around the USA.

42:36And he said to everyone, you know, take take two minutes and speak to the person next to you.

42:41What is it that makes this so good?

42:44Yeah.

42:44And the thing it was was all you had on the chart was the swells and like a white outline of the USA.

42:52But because it was white you couldn't see it and the tiny tight on the bottom right.

42:55It's the complete lack of design that was the design.

42:58Yes.

42:59And that's so important and th that can be so powerful and the power of white space and thinking about those sort of small design choices is so important when you're

43:06thinking about like letting a visualization breathe and allowing people to not feel overwhelmed by what they're consuming.

43:14Exactly.

43:16I think we have it.

43:17I think we've um I think we've covered this uh in as much detail as we possibly can.

43:22Uh we're really interested in people's thoughts there.

43:24I think this discussion is one that can be had many times over

43:28um you know in different places and you probably arrive at completely different um sort of understandings.

43:34Um maybe you think our summary was too brief.

43:37Maybe you've got a great data source.

43:39Let us know.

43:40by all means comment on Twitter, send us send us your emails and feedback in as always.

43:44We'd really appreciate that.

43:47As we highlighted at the beginning of the show, we're going to spend some time now talking about how we record the podcast.

43:53Um but until the next episode, if you're gonna hop off now, or catch you catch you in the next one.

43:59Cheers.

44:00Take it easy if you're going.

44:02Okay, so I'm back.

44:04Uh we actually recorded the section about podcasting and it took uh 15 minutes.

44:08So what we've done is we've split it out into a separate

44:11episode a bonus episode which will release at the same time as this episode so if you still listening out for that just look for the next podcast and you can listen to it there