“Every decision that we make, whether we’re thinking about climate change, poverty, you know happiness, agriculture… how do we feed everyone, how do we clothe them, all of that is underpinned by data and mathematics.
“having a good understanding of data and a good understanding of you know the maths that goes alongside it, that is where our power lies. As a culture, as a society, in how we’re going to move forward the best possible way. And so the more people that can understand data, understand its importance, and understand how to use it, the better for everyone.”
Transcript
Linda: Thank you for joining us for another episode of Make Me Data Literate. I’m really excited about this one. It’s going to be a lot of fun and we are interviewing today. Dr. Melissa Humphries, welcome. Thank you for joining us.Â
Melissa: Thank you so much for having me. I’m so excited to be here.
Linda: It’s going to be a great fun. Can you tell us who are you and what do you do?
Melissa: Okay, so I am Melissa Humphries. I am a statistician at the University of Adelaide and I say that I’m a statistician, not lightly, because I have had a bit of a crisis of like, of identity recently, where I’m like, do I call myself a statistician? Do I call myself a data scientist? Am I a machine learning expert? Do I say that I am in AI? But kind of actually, statistician is like a bit of all of that. It’s a bit of maths. It’s a bit of all of that. And so I’m sticking, I’m sticking with statistician, whether that is the right thing to do or not. I do not know, but that is, that is what I do. I’m also a superstar of stem stem, which is quite exciting. So yeah, at the University of Adelaide and the School of Maths and yeah, part of the superstar of stem program, which is also very cool.
Linda: That’s awesome. And I relate so hard to that identity crisis, because, you know, when people say, well, you know, what do you do? Like, my training is as a computer scientist, and I still identify as a computer scientist. I’m not really a data scientist. I teach data science stuff, but, you know, I don’t do hardcore data science. So I don’t feel right calling myself a data scientist. And I’m not really a statistician. I’m definitely not a statistician at all.
Melissa: Yeah. It’s so hard, because people label you then when you say, like, I’m a computer scientist, or I’m a statistician, they go, oh, boring. Well, if you’re a statistician for a statistician, they do. And you’re like, actually, it’s kind of not like, so I’m on a mission to like, elevate the title a little bit and be like, actually, we’re the ones that make the cool stuff. You know?
Linda: I love that. I love that so much. That is, I mean, it’s part of why I do what I do, right? This is, this is, it’s a tool, it’s a weapon we can use to make really powerful change. And it’s not boring at all when you actually see the purpose of it. And it’s not about, you know, picking white balls and black balls out of urn, and which is the way I was taught statistics and couldn’t see the point at all. I’m laughing because Mel asked me at the start before we started recording, is it okay if I go a little off track? And I see that we already have, which I’m, I just love, you know, it’s, it’s where the interesting stuff happens. Okay, so what did you have to learn to do your work? Was there anything missing from your formal education that you kind of had to pick up as you go? Or was it all kind of built into your degree?
Melissa: Oh yeah. So, yeah, so that’s, that’s a really interesting question because, you know, like I’m relatively new at this compared to everybody. Like I’ve had careers before I did this, right? So when I say, did I have everything I needed out of high school? Well, probably not because I thought that maths was the most boring thing in the world. And I never wanted to do it again. I didn’t do it. And I became a chef and I did lots of other crazy things. Before I came back to uni to do psychology actually, and then kind of got sidetracked into this. Yeah. So I had enrolled in a Bachelor of Arts and then they switched me to a Bachelor of Science because I wasn’t down with the English thing. Anyway, I did my degree at the University of Tasmania, which is an absolutely amazing institution. And the lecturers that I had were in mathematics were so inspiring and so amazing that it didn’t take long for me to realise that actually the mathematics was kind of the coolest part of what I was doing.
And although my PhD was still in mathematical psychology and looking at cognition and modelling that, it was a statistics PhD in the end. So my undergrad gave me the skills that I needed to be able to go out and do great statistics, you know, at a level of what you would probably call data science now, a little bit of machine learning, but you know, some really solid data science with really solid communication. And I think that was one of the really unique things around the University of Tasmania degree at the time was that they focused a lot on how you take this information and communicate it in a really unbiased and clear way that isn’t going to mislead people into making decisions they shouldn’t be making. And so that aspect of my degree was really robust. They focused a lot on the kind of philosophy around why do we make the decisions that we make and how do we make them and how does what we’re thinking kind of impact, you know, how we collect our data and then how we analyse it. So like actually quite apart from the mathematics, just the way we think can impact the results that we get. And that was kind of built into my degree, which was beautiful.
Linda: That’s fantastic because I never hear that. And it’s a big part of the work that I do and what I’m trying to build into the school-based education that we do. But I don’t hear it from people about formal education. It’s really rare.
Melissa: Yeah, I had an amazing lecturer – Simon Wotherspoon. That’s a shout out to him wherever he may be right now doing incredible things. Who really kind of focused on a big part of his lectures and teachings were kind of targeted in that way. And so then I sort of built that into the way that I teach beyond that as well. And that is bringing something that like you say, you just don’t hear about that in a lot of other places. And yet it’s absolutely integral to how we work with our data and how we interpret it and use it. So I’m trying to build that into the pathways that we’re building here at the University of Adelaide and into the new university that we’re merging at the moment.
Linda: That’s so awesome. So do you feel like you came into this equipped to do what you do? Do you feel like you had to kind of figure things out for yourself or was it all like you had it at your fingertips?
Melissa: Yes. So the maths and the stats, yes, I had what I needed. I had the technical skills that I needed. And as somebody who was a bit mature age who had already worked in business management and organised people and trained people. And I also had a set of skills that were able to help me to teach better and communicate better. And were those skills explicitly taught in my degree? No. You know, those kinds of like communication and leadership sort of things are not as explicitly woven into. So a lot of the successes that I’ve had in the role that I have now are a merged kind of like, you know, thanks to a merged component of skills that are partially from my degree and partially from all of my other roles that I have before I came back.
Linda: That’s really interesting to me as well because we treat, often treat those communication based people based skills as kind of you either have them or you don’t. We can’t teach them, which is rubbish. You can absolutely teach them. And there are so many things you can learn that, you know, make so much difference to how you handle a situation, how you work with people. And there have been situations in my work history where I’m like, oh, if I knew then what I know now. I would have handled that so differently. And that could absolutely have been taught, but it was never a part of any of my training at all, not my PhD, not my undergrad. I actually not even my teaching degree really. I did learn a lot from my PhD supervisor, Damian Conway, because he’s an extraordinary person and has incredible people skills and communication skills. But that wasn’t… that was just kind of … I was lucky that I had such an amazing supervisor. It certainly wasn’t part of the part of the process.
Melissa: Yes.
Linda: Is there anything that you wish everyone knew about data? One thing that would, you know, like make everything so much better if everybody understood this this core idea?
Melissa: Yes, I mean, like I have like a dozen of them. Okay, then the number one thing, the most important thing about data that, like if everybody could just understand this, then that would lead to all of the other things that you need to understand. And that is that it’s not intuitive. What you think you should do is probably not what you should do. Like actually, you know, data, the way we manage it, the way we think about it, the way probability, all of that kind of stuff that kind of underpins all the decisions that we’re going to make. It’s not something that you can just kind of go, well, this makes sense to me. So I think that’s going to work.
It’s actually quite a complex thing, even the, you know, and not complex in a way that it’s going to take you years and years and years to learn. I’m not trying to say that. This is something that is achievable that everybody can become literate in. And what I’m trying to say, I guess, is don’t just jump in with the first thing that you think, because often that first thing is not quite right. So just take a step back before you start and maybe read a little bit, ask some questions of some people, you know, and just double check that your first intuition is the right thing to do, because often it’s kind of not.
Linda: I love that. That’s, that’s, um, it’s kind of fundamental to critically evaluating your own work and process, isn’t it, that you stop and go, uh, hang on a minute. What else could this be? Or how else could I approach this? So is this actually the right thing to do? The, um, I have an absolute kind of hardwired tendency to leap without looking. Um, and so I’ve had to really, you know, force myself.
Melissa: It’s a skill.
Linda: You know, it’s, it is. Yeah. It’s another thing that can be taught is to go, wait, before you leap. Not saying you can’t leap, but before you leap. How about you just,
Melissa: just pause.
Linda: Yeah.
Melissa: Yeah. I love that you used the word critical evaluation. So that’s actually the name of the course that I developed here is critical evaluation in data science, because that is actually the fundamental core principle, right? Both, both when you’re doing something for yourself, but also when you’re looking at something that somebody else has produced, like when you read a newspaper article, it is designed to lead you to a specific outcome, right? And if you can pause and critically evaluate, hang on a minute, where did that number come from? What is that percentage that they’ve given here actually mean? You know, like does the statement that they’ve made following that number properly follow, you know, then often you get to a point where you’re like, maybe these statements are a little bit overinflated or maybe it doesn’t quite follow that way. Or, or I actually just have questions that I need to fill out now to make sure that I am on board with believing, you know, the story that they’re told.
Linda: 100%. Yeah, that’s that that idea that we tend to bend at the knees when we see a graph or get given some statistics and go, oh, that must be right. You know, like, oh, it’s numbers, must be serious. And, and to actually pause and go, hang on a minute, is that justified, particularly with correlation and causation, because that, again, that tendency to just accept it and go, well, obviously, this happened after that. Therefore, or, you know, these these two things kind of move together, therefore, one causes the other. It’s like, that’s, yeah, do we have evidence for that? Yes, always the question you got to come back to.
Melissa: Exactly. And even more than that, a lot of the time, those graphics and those pictures that they’ve given you are designed to try and encourage you to think a specific way without even looking at the numbers. There was an example of this just a couple of days ago in a newspaper where they put like the pie chart with the percentages of, of nuclear energy, like who supports nuclear energy. And there was like, you know, 45 or 40% says yes, and 38% says no, and 22 are like, unsure. But they put the like 22 in the 33 bucket and put the 38 in the 22. Anyway, the effect was, there was a big piece of pie with a little number in it and a small piece of pie with a big number in it. And most people won’t look at the numbers, they’ll just look at the pie, you know, they’ll just look at the color bits and go, wow, a lot of people are unsure. You’re like, actually, they’ve, they’ve literally got the figure wrong. And it’s now misleading you to think that there is more uncertainty around this topic than there actually is, you know. And so, yeah, that critical evaluation piece, right?
Yeah, I know, I know, it’s crazy. It’s absolutely crazy.
Linda: It happens all the time, though. There was, there was a whole set of graphs that appeared around one federal election years ago, where it was like, they use the same graph every day and the numbers were different. I was like, what? Yes. Who, who is approving this stuff?
Melissa: I know. But you know, sometimes, sometimes it’s more subtle than that. So there were these graphics that were put out in the United States during COVID, right? And they had, they had circles that kind of grew in size, as, as like the hotspot grew, right? Like as things got more. The problem was, is that so they would publish, you know, one of these pictures every week or, you know, every few days so that people could see how they were tracking across the states. The problem was that actually the legend on the side of the graphic changed every time that they had. So, so you could have the same size circle that today represented 100,000 people and tomorrow that same size circle now represents 300,000 people. And so people are looking at this graphic and thinking that their state or region is staying the same or it’s improving when actually it was dramatically getting worse. It’s just that that, they’d changed the way that the graphic was being presented, you know.
Linda: What really gets me about that is that often the software does that automatically because it autoscales and it, and I’m just like, we’re setting ourselves up for this.
Melissa: We absolutely are. And there in that is the critical evaluation of the self, right? Like, I’m going to put this graphic together. My intuition is I should just go with what the base says, pause and think about is the base setting that is, you know, generating these graphics and is that appropriate for what I’m putting together now? You know, I mean, the classic is colors, right? The default graphic is usually red, blue, green. Please don’t use red, blue, green. Like, 13% of men are red, green, colorblind. They’re not going to be able to see what you have just put on your graphic. Like, you know, change the default and make it appropriate, you know?
Linda: Yeah. I’m so into this conversation. I’ve completely lost track of my questions. What are the worst data mistakes that you’ve seen? And this is, this is mistakes rather than deliberate misuse, although it can be tricky to tell the difference sometimes.
Melissa: Yeah, yeah, it can be tricky. I mean, the worst data mistakes that I have seen always come back to that philosophy of how we collected the data and what we intended to do at the time. So, you know, there was one a few years ago where they decided that they could, they could decide, they could tell machine learning AI could, could tell whether you were going to be a criminal or not based on, based on your facial, based on your facial features, right?
Linda: Oh, phrenology again!
Melissa: And what they found was it was an angle between the edge of your nose and the edge of your mouth that dictated whether you were more likely to be a criminal or not. The problem is, is that the data set they trained it on were mugshots of criminals. And, you know, your little like work card that has like your photo on it, your work profile picture. And so literally, they had just trained a smile detector, because all the criminals were frowning and all of the work people were smiling. And so of course, that angle between your nose and the corner of your mouth had changed. That is, that is not, I don’t think anybody was trying to be sinister in that. It’s just that when they collected their data, they really didn’t think deeply around, or enough, perhaps, around what the flow on effects of how they chose their data and how they sampled their data was going to have on the, on the, you know, end result.
Linda: Yeah. Yep.
Melissa: If you don’t have the right data, and you’ve built in some kind of inherent bias in it, whether you intended to or not, your output is also going to be biased. And that is where your mistakes come from. You know, a lot of these unintentional mistakes from like just really poorly collected data.
Linda: Yeah. Yeah. It’s so easy to see in hindsight, but you can kind of see how this happens in the process of the research, you know, you’re like, really, really excited about it. You’re like, Oh, this is so interesting. And you just don’t necessarily evaluate, … you know, the number of, number of data sets that have been trained on white men because they were, you know, it was like photos taken around the office and it happened to be only white men in that office. I mean, technology.
Melissa: Yes. Yes. Yeah. Exactly.
Linda: But like, these are the people who understand, or these are the people who are representatives. Like, maybe there’s some people missing.
Melissa: Yes. Exactly. And this is, and I mean, you know, part of that, you can understand how this happens. You, there’s limited funding, there’s limited time, you’ve got to be able to access this data. People have to give permission for their data to be accessed. You know…
Linda: that doesn’t seem to be a thing anymore. I’d like that to be a thing, but it doesn’t seem to be the case. That whole permission thing, we seem to be glossing over that.
Melissa: I know you. I was at a thing the other day and they said that when you walk into a supermarket, there’s actually like fine print that says you can use our free Wi-Fi. But even if you don’t, we will just like tap into your phone anyway, unless you opt out. I was actually kind of shocked that just walking into a shopping center or like a, you know, like a mall or, you know, could actually, they could just access that. That is your permission. You’ve given them permission. Double check next time you go somewhere. Like, I haven’t been anywhere since, so I haven’t checked. Double check that fine print.
Linda: And I bet the opting out is by writing a letter on paper and posting it to an address in the calendar or something. Oh boy. That’s kind of a good segue into, have you ever seen data deliberately misused?
Melissa: Um, I mean, yeah, we’ve all, I mean, we have all seen data deliberately misused. We see it all the time in the news and in politics. And, you know, sometimes it is very small and it doesn’t make much difference. Sometimes it is absolutely overinflated to try and lead people into thinking something that isn’t, isn’t quite right. And so I think actually this is, this is part of the reason why I love your podcast. And I love the idea of like, you know, trying to help people become a little bit more data literate. And that is not saying that people are data illiterate. It just means that there is like a level of reflection that people need, to make sure that you are not being manipulated into thinking things or feeling things that are actually not quite right, you know.
And so making sure that you are critical of, you know, I think that there’s this classic example of amount of money spent on food stamps in the US. And there was this huge story that was like, you know, there’s like a million, a million dollars is, you know, or something bigger than that, is actually people are scamming it out of the system every year. And this is horrific. And you know, and it’s a big number of million dollars or like 1.7 million or something. But actually, when you look at how much money is given over to food stamps every year, which is like billion, you know, or hundreds of millions of dollars. And then you look at kind of the rate at which this kind of scamming is happening. It’s actually way, way less than even at a supermarket, you know. And so, so, you know, presenting the data in a way that is, it is designed to make you go, oh, no, that’s horrific. If you, if you see a stat like that, and it gives you that kind of visceral reaction, then that is an absolute red flag that you should probably look a little bit deeper into that statistic and make sure that it is genuinely as shocking as you think it is, you know.
Linda: Yeah. It’s, it’s always interesting to see those kinds of clickbait data studies actually broken down and go, well, you know, hang on a minute, what, what is that as a proportion, you know, like, what is that per capita? What is, you know, that sort of thing, you know, you can, you can put things wildly out of context and go, well, you know, how many deaths have we had according to this? Well, we’ve had so few compared to the US, yes, but we are like a tenth of the size, you know, maybe
Melissa: exactly.
Linda: Let’s look at it per capita, you know, that, that sort of thing can be so significant. This seems like a good time to talk about the study that you and Lewis Mitchell put out, came reported in the conversation just a few days ago, and I was so excited to see it because there’s been immense drama in the media, particularly through some books that have come out and stuff and people trying to build up this hysteria around the idea that social media is making our children depressed. Rather than say looking at the world around them and, you know, being, we’ve got some problems here, maybe that’s an issue.
Melissa: Yeah.
Linda: it’s very frustrating because there’s a lot of, well, you know, mobile phones appeared and then kids started getting depressed and you’re like, this is, is it correlation versus causation thing?
Melissa: Yes, yes.
Linda: Does this happen at the same time? Doesn’t mean it’s one cause the other. And finally, you and Lewis actually put some data around it. Tell us, tell us about that study and how it came about and what you found.
Melissa: Yeah, so we, we were, we were interested in looking into this relationship between social media use and sadness. And as we started digging into the literature, I mean, I was kind of expecting, I was part of me was expecting to see that, you know, social media use goes up, sadness goes up, or, you know, because people talk about that a lot.
But a lot of the recent research has dug into that relationship much deeper than that kind of superficial, you know, you know, computer use goes up, happiness goes down, kind of findings that we had a few years ago. And they’re finding that the pattern is a lot more complex than just, you know, social media use is causing this. And in fact, there is some evidence to suggest that it’s the other way around, that if your mental health is low, and that if you are feeling depressed or experiencing, you know, other hardship, then that means that you tend to use social media more, right, rather than it being the other way around.
So some of the things that I found really interesting in the research that we found was that the patterns are different for different age groups. So for older people, if they were able to connect with people on social media outside of their family, so, you know, groups or, you know, different teams or whatever, their happiness increased. It’s like their connections and their sense of belonging and their sense of acceptance increased making them happier, right. Whereas for young people, if they were able to interact with their family online as well through social media, then their happiness increased.
So, you know, different ages and different groups of people need different things from social media, but it all revolves around this feeling of like connection, community, support. And the reality of social media these days is that we do get that from the social media that we engage with.
Young people talk about, you know, feeling like they have, they’re empowered to, you know, be activists and, you know, contribute to the conversation around what they want to see our country and our world doing, that they can’t contribute to formally yet because they’re not old enough to vote, right. And so it’s a really important space for our young people to be able to have opinions and learn and grow and connect.
And that’s not to say that there aren’t other things on social media that are harmful as well, but this is part of the journey that we need to take with our kids, right. We need to be able to support them to understand how to use it well and look out for their own safety.
Linda: That’s wonderful. I love that. I love the idea that you didn’t find what you expected to start with. I think that’s a sign of good research. You know, when you find what you expect, I think you need to examine it a little more closely to be sure that you didn’t just go looking for what you wanted to find and, you know, and therefore find it in the questions that you ask can be so formative of the results that you get.
Melissa: Do you know something else that we found in the research that I thought was like actually really interesting? that rain – rainfall is correlated with sadder sentiment in your post, right. But not only that, if you’re experiencing rain, the sadness of your posts brings down the happiness level of the people who are connected with you as well who aren’t experiencing rain. So actually there’s this relationship between the kind of emotional content of the post that you’re looking at and that flow on effect can flow to the people who are connected with you. But happy posts are way more effective at affecting the happiness of other people than sad posts are. So you can see like three times as more happiness in the people in your networks if you are like experiencing happy posts than sad ones.
And so I kind of have taken away this like happy makes happy kind of vibe that and you know I’m not saying Marie Kondo all of the sad people out of your social media
Linda: <laughs> that’s harsh
Melissa: but I am saying, you know, I am saying make sure that you have some like some some good and fun stuff to look at as well, right. Like make sure that you’re connecting with some of that happy stuff and sharing your own happy stuff so that you know we can elevate that kind of community network of happiness that is out there.
Linda: That’s beautiful that ties in so well with my own experience because I realized I so a lot of the social media that I’m on I’m on because my kids are on and I was like okay well I have to you know just see what’s going on make sure I know what’s you know what they’re into. And so I’m on TikTok and reluctantly because I hate I hate video as a form of… I like to read I’m I’m I’m text driven I’m all about reading and I found that TikTok was kind of sending me down these really weird and disturbing rabbit holes. there was a while where it was sending me down this rabbit hole of people trying to break into your hotel room and I was getting video after video of this you know people putting things under the door and kind of doing it the lock and stuff that actually doesn’t work in any hotel room I’ve been in in the last 15 years anyway. But also like what? why was I getting this stuff> and so I’ve got to the point where I flick past anything that’s not a cute animal or David Tent wearing a trans supporting t-shirt or you know or snippets from my favorite shows and that’s like that’s it that’s that’s pretty much all I see, and and some comedians and I’m very careful about which comedians.
Melissa: Yes.
Linda: So I’m pretty much queer comedians only unless I’ve got you know really good evidence of this this one this one is safe. And and I’m so much happier now you know if I’m feeling sad I just go and watch cute cat videos and yeah it upsets my cat so maybe there’s a bit of a problem there because she’s like what are those cat noises what is going on there are interlopers in my house.
Melissa: So funny.
Linda: But it makes me much happier you know because I’m making sure that the stuff that I see when I’m idly scrolling social media which I usually do when I’m exhausted or sad. Making sure the stuff that I see is healthy for me and not taking me down some paranoid horrific rabbit hole.
Melissa: Yes yes and see in that again we come back to that kind of self-critical reflection right of like if I’m feeling sad how do I now make sure that my social media is supporting me in the way that I need it to. So one of the quotes out of the thing is you know are you something like a are you using social media or is social media using you.
Linda: I love that line.
Melissa: Yeah and I just I just think you know if you can keep checking in with yourself occasionally to make sure that you are actually using social media and that it is serving the purpose that you need it to that it’s providing you your connections or your happiness or your outlet or your you know the information that you need whatever it is make sure that that is actually serving its purpose and that you haven’t flipped over into that other world because it’s so easy to go down rabbit holes. it’s so easy to find yourself in a in a spot where things are not actually serving the purpose that you wanted it to but it’s also possible to change that and notice that and say okay I’m going to stop following this page or I’m going to you know flick through these reels so that I don’t get any more videos about this kind of stuff or you know take action to make sure it’s giving you what you need.
Linda: Yeah that’s I love that but I think this is going to be a separate blog post just this part of the conversation because it’s such an important message. To come back to the the questions again what is the first question you ask when you look at graphs in the media?
Melissa: The first thing I do is I look at their axes because often they do weird things with axes to make sure that they’re not that they’re not to try and amplify you know effects and things so you know I’ve seen all sorts of things you know the the simplest thing is where they don’t start kind of an axis at zero they started like a million and so it looks as though there’s this massive difference between the like bars that you’re seeing and and there isn’t actually. it’s just that they’ve changed it, but they also do weird stuff where they’ll like have you know big gap between zero and ten and then a smaller gap between ten and a hundred and then a smaller gap between like a hundred and a million and so it looks as though something’s flattening out or things are not changing very much when actually the opposite is true and so axes are my are my first thing.
The second thing that I do is actually check that the numbers that they’ve presented match the picture that they’ve presented because like I said before often the pie chart the actual angles don’t match the numbers or or you know one bar will say 20 and one bar will say 2000 but they’re pictured like super close to one another you know and it just doesn’t you know
and to be honest with you if I see a graphic that is made out of animals or fruits or you know I just don’t even look at it because I know that it’s going to be wildly misleading and that somebody has they call this a duck and it’s it’s um these graphics called ducks which are designed to be pretty first and informative second and it’s called a duck because there is a house I think it’s in the UK that somebody has built like a duck and it is the most impractical house you’ve ever seen in your life but it does look like a duck so it clearly serves the purpose of being the duck but just probably not that much the house right and so and so when you see these kind of graphics that have like oil barrels or stacks of coins or books or pineapples or you know you know that they’ve tried to make something pretty um yeah and that it’s probably not going to actually convey the message that it’s meant to convey properly
Linda: yeah yeah that’s a real challenge I used to um when I was teaching my year 10s a data science course um and and trying to get them to build what we were learning into their science projects, their extended experimental investigations they would come to me with sort of what wanting to do a graph and and trying to make it interesting because that was part of what we were focusing on was maybe not just a graph let’s see how we can make it really compelling and tell the story that the data actually tells uh really effectively and it’s not easy you know to make if you’re using fish and changing the size of the fish to to represent how many of that type of fish there were. uh how do you you know how do you do scale is it area?
we’re not that great at detecting area of sort of separate objects like that visually so you know is area is a position on the screen is that you know how do you do that? and and um it’s very easy to get that wrong I think it’s very easy to be misleading even when you’re sort of mathematically correct you know well this is actually twice the size of that one yes but it doesn’t look it and that it knows that you have to deal with perception not just maths
Melissa: yes and this is where I mean graphics is graphics is and and you know data visualization it’s like a field in and of itself right and this is where it becomes really complex it’s like this wonderful like collaboration between psychology and mathematics in that the way that we perceive things, the illusions that we see, can actually make us believe things that aren’t even true on the page in front of us. like you say it can be mathematically correct but actually appear to be something entirely different just because of the way that we process information. you know. the way that our brains have been trained um and so I mean and people have have varying different ways of trying to trying to you know attack this problem and and there are some incredible people out there that make incredible graphics that are beautiful and and intriguing.
my advice has always been to the people that are coming through that actually the point of your graphic, the point you’re trying to make? that point, that message should be impactful enough that that message is the thing that is going to make people go: wow! really? you shouldn’t need a fish. you shouldn’t need a pineapple. you know the message itself should be – and if it’s not then maybe you don’t need a graphic maybe that can just go in the text somewhere. do you know what I mean? and so um I’m not saying that nobody should use fish or pineapples like go for it and try try your best to do something really funky you know but um
I do think that for a lot of the kind of statistics and things that we’re trying to convey especially the stuff that we see in newspapers and magazines and you know those little those little dashboards that come up online and you know um having just some really clear and precise numbers or or like you know bars or something that is just really easy to see is often enough right, to go whoa that difference is huge or or I didn’t realize 55 percent of people felt that way, or you know like um so be true to your message I guess is the thing that I try and tell my students about that stuff
Linda: yeah yeah and always come back to sort of the original graph to to see you know is the is the message the same is the message really the same between this exciting picture and this actual you know mathematical graph
Melissa: and you know you can always like give it to your mum, your dad, your sister your housemate you know like your colleague and say what does this graphic tell you and if they go oh um oh I think it says I don’t oh does it say like… if you have to explain to them verbally what they should be seeing in the graphic that you’ve just then the graphic needs some more work
Linda: yeah 100 percent it’s like um usability if you have to put a label under to tell people how it works then it’s not usable you know if you have to put a sign on the door that says push you have failed in designing your door. you know just don’t put a handle on it put a plate so just you push because pushing is all you can do then you’ve won you know like yes. people people don’t read you know, breaks my heart but people don’t read by default. so if you have to put a label on it you’ve lost them.
Melissa: yeah
Linda: this this has been great I think we’re gonna have to do this this again um what excites you about data
Melissa: ah everything! like I mean honestly there were lots of different things I could have done for my phd, like a lot of different directions that I could have gone in, and I have chosen data because data is data is the future of our entire planet like everything that we do every decision that we make whether we’re thinking about climate change poverty you know happiness agriculture you know how do we feed everyone how do we clothe them how do we how do we make so all of that is underpinned by data and mathematics all of it every single bit even our art even our art and our music and our like all of it has like data threaded through it these days all of it and I mean for me selfishly that means I get to like dabble in everything right I get to like still still be attached to like a whole bunch of different things that are really exciting and make me happy um I don’t have to just just play in one sandpit for the rest of my life you know.
but it does mean that having a good understanding of data and a good understanding of you know the maths that goes alongside it that is where our power lies as you know as a culture as a society in how we’re going to move forward the best possible way and so the more people that can understand data understand its importance and understand how to use it the better for everyone
Linda: what a what a perfect note to end on that’s and I didn’t prime Melissa to say that, that was all her own work! but that’s I mean that’s what it all comes down to isn’t it that’s the core of it and we do people a real disservice when we scare them away from data with the way we present it and the way we talk about it and the way we teach it. thank you so much this has been fabulous, I really enjoyed this conversation!
Melissa: yeah me too this has been absolutely amazing I’m so grateful to be here
