When to go interactive

 

So I want to discuss when you might think about ticking things and going interactive. And listen, the spoiler alert on this one is, my answer is kind of, always, sort of. If you can, if you have the time and the budget to make interactive things, you should. So that's the short answer, but here's the longer answer. I like to start off this way. 

How did you learn to count? When I was a little kid, I learned using these, they're called Montessori counting beads. And so when I was three years old, someone handed me three of those little beads in the upper-left-hand corner and said, that's three. And that's not that weird. Everybody learns to count that way probably with blocks or something like that. But then they handed me this single-wired together object in the far-upper-right and said, that's 10. 

Then they handed me the flat object in front of those in the lower-right-hand corner, which has 100 beads all wired together into a single object. So when I was three, I think I understood that 10 times 10 was 100. 

Okay? So that's very tangible. This physical experience of learning numbers, made numbers very tangible to me, and interactivity makes your data tangible to your audience. 

They're actively participating, they're touching and affecting and filtering and sorting the data. So they have the experience really more deeply than if they're just passively watching. So that's the first argument. The second argument is this. Everyone's a narcissist. 

Okay? And I mean that in the nicest way possible. When we're consuming content, if it has nothing to do with us, then we can't relate. 

Okay? So we want to make our content relatable, make it as much about our audience as we possibly can. So the example I like to give is this. If you are from deep in the Amazon, you might not enjoy watching a hockey game on TV. 

Maybe at first, it's interesting, right? Like, hey, why is that stuff so slippery? And what's that black disc everybody's chasing around? But after a little while, it might be boring because you don't understand the rules. 

Everything is so foreign, there's not enough about me to keep my interest. So as an example of a project that I created where the interactivity really helps bring this data to life, it's something created for the Boston Public School system for their 10-Year Master Planning. 

And so we have all the schools in the Boston Public School System, and you can go in, you can see a map of the schools and where they exist in the city. And each school has an individual profile. 

So I can find a particular school by name and click into it and see its data and have measures on different scores, et cetera, et cetera. And this is all well and good. It's good data. But this is essentially sort of replicating what was in the printed report, which was a fine document that had a lot of good stuff in it. 

But in addition, we have the ability to go in and analyze the data in more depth and look at it from a different direction. So I get these essential distributions to look at all the different data points as far as accessibility goes, where how many schools fall into different categories on accessibility or the quality of their heating distribution systems, et cetera. And I can highlight a specific school, make it about me. Where does Bates School live? Aha. I see which category they're in all along the way. And I can of course filter by different things as well. The point is, as a parent, as an administrator, seeing the overall data is nice, hearing what you're going to do with this data as far as the city is concerned, is helpful and interesting. 

But I want to see where my school is. I want to see how this is going to affect my family and my kids. And so if I can make it about me, the way I like to think about it is it allows you to turn your data story into your audiences data story, which is very powerful. And a related issue by the way, is credibility. So in this particular example, parents and administrators might question the logic. If say the outcome was, the recommendation was to close a school. Why? 

Well, I can go through and dig through the data and see how my school compares to all the other schools in context. And another example in terms of credibility is transparency. And so one of the features built into this tool, is that the map, as I zoom in, more information is revealed. Specifically the plan. 

Okay? And I see not only the layout of each individual school building, but also the layout of the land that the city owns around the building. Owns or, and or leases, I think. And so, a long story short, if the argument was to say, close this school that's an orange here, or that, you know, hey, we can't expand it, so we have to do something somewhere else. So I can see that the city doesn't own enough land around that building to do much else versus this red building where there's a ton of land, plenty of room to expand, et cetera, et cetera. Now that transparency might enable parents to make an argument such as, okay, well, why don't we take that orange school and build a version of it on that land that the red school occupies? Hey, that's the thing with transparency, right? You give your audience the ability to make arguments with your data, which is not a bad thing overall. 

So interactivity really is about letting your users explore data, which enables things like making it tangible, making your data story into my data story, and increasing transparency and credibility. Simply put, you should go interactive when you can. Allow your audience to find their own Aha moments in your data, expand the reach and the depth of your content. 

And by the way, get past the editorial constraints of a non-interactive thing. Even a deep, deep report has stuff left out of it, has sorting and filtering decisions predetermined. Allow your audience to get past that and find their own pathways through your data.

How to think interactively

Selecting transcript lines in this section will navigate to timestamp in the video

- So, you've decided to create an interactive experience, which is good, as I said, I think you should pretty much do it if you can. But if you're new to interactive, it can seem like a new world. You had to think about being interactive versus creating static experiences if that's what you're used to doing. 

But, you probably use interactive technology all day, every day, right? Your phone, you're on the web all the time, you're on Facebook, et cetera. So the leap really isn't that huge if you think about it as a user. In this movie, we're going to cover a few ways to take the information you have and sort of flip it on its head a little bit, think interactively, and create productive experiences for your audiences. The first rule in thinking interactively is don't think interactively. 

So it sounds a little counterintuitive, but I recommend that you don't think about features and buttons and actions at first. The first thing you want to do is look at your data, think about the story that you want to tell, think about your audience, all the things we talked about at the beginning of this course. 

So thinking interactively really is sort of like a microcosm of this entire course. It's a microcosm of the whole design experience. We're just sort of taking it to another level. Creating ways to expose more knowledge for your users through touching and clicking and rolling over. Really want to go back to the basics in thinking about your design in general, before you think about features. 

Good place to start is to think about, when I look at the data that I have, what's missing? What's not there that I wish I could see? And I don't mean what's literally missing, but what's not available to me? What's not easy to parse? This is where I usually start to think about features to add to an interactive experience to expose what's missing. 

In addition to what's missing, what can't I see? It may be there, but it's sort of like a forest versus the trees issue, can I bring more to the surface and show both the forest view and the trees view for my users? Interactivity is great at shedding light on what's hidden or occluded in your data. So here's an example of interactive experience that I created, and I'm going to start off by admitting to you that I did this in exactly the wrong way, I did it completely backwards. I followed none of the advice of the rest of this course. I didn't start thinking it through, I didn't plan it, I didn't think about my audience, I didn't design it, I didn't go through the whole process. I literally started coding first. 

Now, that's the wrong way to do this, except that it's a great way to explain how to think interactively, 'cause with a very pure experience of looking at the data I had, thinking of features and how to release the story from that perspective. So, I don't recommend this, but I'm going to talk about that way to help you think about interactivity. So what we're looking at is one dot for every single team in the NHL, the National Hockey League, and every year of play. So as you can see, a couple labels here, the 1929, 1930 Bruins are here. We're looking over here at the 1995, '96 Detroit Red Wings. So you have about 1,300 data points in this one bubble chart. 

Now when I first created this, I'm going to scroll down a little bit so you can see, there were no filters at the top, and if you can imagine, there were no labels and no colors and no legend, and nothing, I just had an axis and a bunch of dots on the screen. This is the data that I had available to me. So, as I mentioned, there are 1,300 or so rows of data, one for every season for every team. And back up to the top here you can see I had a code for the team, season, how many games they played, wins, losses, ties, points that they got for the season. 

And just for point of reference, you get two points for a win, one point for an overtime win or tie. And points percentage is how I'm measuring quote on quote, the best team in hockey, and that is the total points of team one divided by the total that they could have won if they'd won every single game. And we also have a couple more data points here, one of which is strength of schedule, which is, were they playing a lot of hard teams that year, or a lot of easy teams that year? 

And then some other data that I'm mostly not using. But there's a lot of data in here to look at. So, again, back to the scatter plot. As I mentioned, I had a lot of data to work with. I had a simple scatter plot that had a lot of dots on it, but my first question was, what's missing? What's hidden? What might I want to know if I'm a user? And so, first thing first, what do the dots mean as a user, if I'm just using a bunch of gray dots? 

So the first thing I added were these labels, and I decided of course not to label every single dot, 'cause that would have been insanity. I just labeled the best and the worst, right? Red and green, best, worst labels. By the way, side note, I double-checked and I used a color blind friendly red and green, these will both work well for all different types of color blindness. So, now that I have the dots labeled and I know that the best team ever was the 1929, 1930 Bruins, which of course made me happy, being from Boston. 

And the worst team ever, 1974, '75 Washington Capitals. The next thing I might ask myself is, okay, I know the best and the worst, but what about all the rest of the teams? Okay, so thinking interactively, of course it was logical to then add rollovers. So, as I interact with a dot, I can now roll over any dot and find out the details of these teams. And so I go down here and I see these two dots that were just above the '74,'75 Capitals, and they're the 1992, '93 Ottawa Senators, and the same year, San Jose Sharks. What happened that year? 

Two teams that really struggled, interesting. So, back to what's missing, what might I want to know as a user, again, driving me towards functionality. I'm looking at the Bruins up there and I'm happy to see that, but I know that my friends in Montreal are going to say to me, Bill, come on, the Bruins are not the best team ever, it's the Montreal Canadiens. They've won 26 Stanley Cups over the years, the Bruins have only won six. And so I said, alright, let's give tools to allow people to filter the data to find what they might find useful. So that's when I started adding these filters you see at the top here. 

So I can go in here and filter by teams. Alright, my guy from Montreal, let's let him sort and find just the Montreal Canadiens and look at them. And yes, see the dotted outline means Stanley Cup winners. And sure enough, there's a lot of Stanley Cup winners in there, he's right. I also might want to say, okay, let's just look at Stanley Cup winners. What does that look like? Okay, by that measure, the '43, '44 Montreal Canadiens were the best ever. Look at this, '37, '38 Chicago Blackhawks. 

They only won 14 games, they really struggled that year as a Stanley Cup winner, that's a very interesting find. I also then said, okay, what about seasons? I might be interested to see different years. So I can go in here, and let's say look at that year where we had those two teams that really struggled. 1992, '93, look at how big that spread is, that's interesting. Or I might want to look at 1996, '97, let's see what we see in there. 

Okay, a much more tight spread of dots, Boston Bruins at the bottom that year. Well, let's look at 2010, 2011, another year the Bruins won the Stanley Cup, all very interesting stuff. Maybe I want to look at decades, maybe I want to look at the 1920s, interesting, a really big spread of numbers, versus the 2000s. Much tighter spread, seems like a more competitive era. It's an interesting way of looking at the data. Finally, and this is where we get to that credibility question, I added a bunch of these, what I called superlatives. 

So, let me look at who won the most games in green, versus who won the least games in red. And this is just sort of the top 50 and bottom 50. And sure enough, the top 50 had very high points percentages, and the bottom 50, for the most part, didn't do so well, with some exceptions, of course because there was some short seasons over the years. But then I said, alright, across these different superlatives, there was one down here called Strength of Schedule. As I mentioned before, that means teams that had either a really easy schedule, that played a bunch of bad teams, or a really hard schedule, they played a bunch of really good teams.