Ben Kraal Transcript

Transcript

1. 1 www.captionslive.com.au | captionslive@outlook.com | 0425 904 255 UX AUSTRALIA Design Research 2020 Day 1 Thursday, 19 March 2020 Captioned by: Gail Kearney & Rebekah Goulevitch
2. 2 BEN KRAAL: This is Ben. Hello. Someone type in the chat if you can hear me. Hooray. All right. So thank you. My name is Ben Kraal. I'm a design researcher. I'm a director at Symplicit and today I'm presenting to you from home in Brisbane where I recognise the Turrbal people north and south of the river as home of the Jagera nation. Design synthesis comes from why we do it, post-it notes, mirror boards or things like that. The reason I want to talk about this is first of all, because I had a great idea for a title and it made people's eyes go crossed, so I might leave it there. What I want to do today is look forward to - I can't see my face. Let's do this. How is that? Lovely to see you all. What I wanted to do today is to look forward and think about what we do when we do design research synthesis but also to look backward. And in this setting explaining our method to us is a weird thing to do. But when we come together at events like this, I think it's really a chance to reconnect and to learn new ways to do this job that is sometimes hard to explain to our partners, let alone our parents. And so looking back, I want to look back at our roots and see how we got here. So the reason that I wanted to give this talk is I was standing in this room that you can see a photo of and the client said to me, "Why do you do it like this? Why does this work". It obviously does work. Like, why? Maybe you've had the same thought, is this the best way to do things, tearing cellophane off post-it notes or a mirror or maybe you want a different way to think about what it is you're doing. Maybe you're like me and diving into that mess is your happy place. Maybe you've moved from looking at post-it notes and looking at transcripts of interviews. John Corko has written in typical detail about his team work with detailed transcripts. There is heaps of software that lets you work that way too. But when you do that, the thing that's happening is it's still the same process in the end and what I want to think about is what are the assumptions that go into that process and what ways of thinking does that process actually help us do. And in doing that maybe we'll figure out
3. 3 how these things actually work. And so just to figure out where these comes from. So here's some spoilers, and I don't believe spoilers ruin a good story and I think the story that holds our methods is a great story. This has fencing, fighting, torture, poison, that is the 'Princess Bride'. This is has bearded potential, naval navigation, death, dying and lies and a 50-year argument about 20,000-year-old ideas about what it means to truly know anything. At some point I will use the word 'epistemology' and I'm not sorry about that. So what does it mean to look back? Well, as far as we can to fight anything we still think of that is still UX. I think if we go back far enough we hit contextual design. To 1998, which is last century, and it's about designing systems that reflect the way customers want to do their work. And so without getting into definitional wars, I think that's a pretty good way to think about what UX is. I like this for two reasons. As an origin, they are specific about using post-it notes to do research synthesis. The other reason is that almost as specific about the foundation of their approach. So I'm going to pick out a few things that they talk about as their philosophical background that informs everything they talk about in contextual design. The first one I'm going to talk about is probably the weirdest and as I go on you will see that's a fairly bold claim. It's a book written by two people, by Terry Winograd, the taller one in the left-hand photo who was an artificial intelligence researcher before he reached closer to the human interaction end of things. In the 1970s he wrote language processing systems that are better than Siri. He was also Larry Page's chief adviser before Larry Page dropped out to start Google. The other person who wrote the book was Fernando Flores, he was in the Government from Chile in the 1970s. To shed light, he was the Marxist forced out in a CIA-staged coup in favour of Pinochet. Fernando was the guy who commissioned project Simonson. It looked like this. It's not too much of a stress that Terry Winograd and Fernando
4. 4 Flores had a communism future. This is about phenomology. A lot of people go what on earth is phenomology. It's an experience of consciousness. If I've got a drink in my hand, I joke to people it's about the philosophical implications of the fact that people are people-shaped. It's understanding how people act in the world. It's about this book about applying that when you're thinking the design of computer systems. This is pretty radical because it's from 1986. But it's also how the way that computer scientists thought about thinking was too limited. So Winograd and Flores say the particular ways about thinking about thinking were built into artificial intelligence systems at the time meant that that system would always be limited and too hard to use. They didn't try and identify ways to fix these systems. They tried to find ways to understand how people used these systems and use that understanding to make the systems better. So what they're ultimately trying to say is computers are a way that people get access to information about the world and because computers are becoming in 19986 a strong way people were getting access to this information, we needed to do better at that to help people make better decisions. So we know this now but this was 1986. This was fairly radical. It' not that computers and cognition is the first computer-human interest action philosophy book, it's not the first to make the connection between phenomenology and design. The next thing is in the 1980s Xerox were like dominant. They were enormous. They had so much money, they had two different research campuses. It meant you bought the smartest researchers in the world and told them to think about things and made money out of that. Photocopiers are getting bigger and bigger and more sophisticated but their customers were complaining they were also harder to use. And so Xerox made even more sophisticated copiers that had a raise of sensors built in. They were smart photocopiers in the 1980s. These sensors in the control software were supposed to help the copier understand what people were doing and give you helpful instructions on how to make
5. 5 copies or collate or how to do double sided and things like that. Unfortunately the instructions weren't helpful and the copiers stayed hard to use. Lucy Suchman was one of several anthropologists Xerox hired. You must have a lot of money if you are a copier company hiring anthropologists. Suchman was interested in the marketing around copiers and how the users used them. She took over the park where they were made and used videos of how people used Xerox. What this means is there's lots and lots of videos floating around of the most world renowned computer scientists failing to use photocopiers at Xerox. Suchman's book is about plans and situated actions about what was happening. She is an anthropology. She drew on all the ways to understand what was floating around in the anthropological situation. It's a way to understand how people use their circumstances to achieve purposeful actions. It's a quote from the book. So she videoed people and closely ana lysed what they were doing. She approached it called conversation analysis. What she found was that the plan that was formulated by the photocopier about what was going on and what the user thought they were doing would get increasingly far apart. And so what's happening is what the photocopier thinks is happening is not what the user is really doing. It's this mismatch that makes the copier hard to use. As you are following the plan, as you're miss interpreting the plan, you get further and further away from what the copier thinks is going on and you fail at your task. Even though the copier was smart, it didn't understand what people were doing. And then that action that people were taking got further out of alignment until the point when someone discovers that there's a problem. You can't figure out what happened because you're three steps down the chain from where the photocopier thought you were. So that's a really good thing to happen to you, to find that you can't get there from here. So this is the mid to late 80s. This in the first wave of AI systems is emerging. And so the book plans and sit actions is also a critique of this point of view that is embedded in these naive AI systems where the idea was you could sense
6. 6 the user's goals ahead of time and then meet those goals or those needs with clever design, saying how that is similar to current users experience is an exercise left to the motivated listener. The next thing that they draw on is a way to think about thinking. And there's a way we think about thinking all the time. We think that thinking happens in our heads and we think that that the way we study thinking often is we put people in a decontextualised room and ask them to solve problems. And it totally works. We've learned heaps about thinking from doing that kind of research, whether it's pure psychology research or often times how sometimes we do research with users. But there's another way to think about thinking. And it starts by seeing that thinking happens in all kinds of places and can't be separated from social and cultural contexts. So an anthropologist, Edwin Hutchins found himself on the bridge of a Navy destroyer and he wrote this book called Cognition in the Wild. It's about how the team of people on the bridge of a boat does navigation. And the thing that he is investigating is not how individuals think but how the team thinks. And so his book is like this introduction to cognitive anthropology for computer scientists or for people who like the metaphor of cognition of thinking as computation. But it's also addressed to anthropologists who at the time Hutchins was writing had given the mind a special place different to culture. What Hutchins was trying to do was for people to see culture as much as thinking as a process that develops solutions to problems. So that's the thing about thinking and culture which Hutchins thinks are bound up in each other, you can't pull them apart. He spent a long time thinking about how navigation happens and there's a lot of calculation in navigation. So he wants to show that the calculation doesn't happen with what is in one person's head and what they are doing to get that thinking outside of their head and how other people use what's outside their head to get that thinking into their head. For example, if you are doing long division or calculus you probably need a piece of paper. The paper doesn't do the thinking but without the paper
7. 7 you can't solve the problem either. So Hutchins would say the property is of the human in interaction with the symbols make the computation. And so what he does, he spends his 400 pages that make this argument that cognition and thinking doesn't happen the way that cognitive and scientists thought it happened. It's not symbol manipulation, it's something more. And the something more is social, emotional and ultimately cultural. So these three books that are part of the philosophical background to the contextual design, you are drawing on these whether you know it or not. What these books have in common are two things. They are arguments that looking at the context through culture and society is just as important to understand technology as understanding technology. You can't separate culture from technology. And equally so they're quite explicitly critiques of what at the time were the dominant ways of thinking about artificial intelligence. So to a certain degree doing design research the way that we think about it and the way we do it is inseparable from the critiques of AI that were happening in the '80s and early '90s. So there's another book and it's a bigger idea behind the way we use post-it notes to do design research and design research synthesis. Even if you have not heard of the book, you have heard of the idea and it's called grounded theory. It was created by two anthropologists, Anselm Strauss and Barney Glaser. They were in the 1960s talking about death and dying. It was controversial work because it became this consensus in the community of sociologists they were part of that had this idea that everything that could be said about death and dying had been said. They, Glaser and Strauss, thought otherwise, otherwise they wouldn't have done this work in hospitals. And they spent lots of time in six different hospitals in the bay area in the late 50s and early 60s. So a lot of their work was just like hanging out. They could spend a whole day sitting at a nursing station listening and eventually they would maybe conduct more formal interviews with people when they had some sort of idea about what they wanted to ask.
8. 8 They wrote this book called ‘Awareness of Dying’. 1965-ish when this came out, death and dying is increasingly institutionalised and it had been for a while. And it was to a greater or lesser degree sort of uncaring. One thing that Glaser and Strauss saw was that doctors and nurses and families were unwilling to talk about what was happening to the dying person. The doctors are focused on the technical aspects of what was going on. Nurses weren't allowed to discuss anything with patients without the doctor's permission and there wasn't the emphasis on the psychological emphasis on nursing that we have today. There is a real effort to keep patients completely in the dark about their own imminent death. And so friends and family were basically recruited to lie to the patient and this meant they weren't able to grieve. As this is happening, the patients have an awareness something is up. They are obviously in hospital and very sick. Glaser and Strauss were deliberately provocative to the doctors and nurses to see if they could get more information about what was happening to them. After a while everyone has to admit that the patient is dying because there's this stack of lies upon lies upon lies. Now, they were trying to do this to preserve some sense of dignity and privacy for the patient but this is causing stress for staff and families at the same time. So eventually what happens is you have this tipping point where everyone just comes out in the open that the patient is basically terminal. And so that's what this book Awareness of Dying is about. To get to this point Glaser and Strauss had to do this qualitative work and they had to be able to show rigorously that what they were saying was happening was real. And the way they did that was they invented grounded theory. And it's called grounded theory because it's grounded in the reality of what's happening. And they are very careful to describe their process of creating their analysis in detail. So that the steps they took to get to their conclusions are just as available for inspection as they are in quantitative methods. One reason people like numbers, as Leisa said earlier this morning, was that they feel that they can inspect them, they can know
9. 9 how the number was arrived at. And so Glaser and Strauss are trying to have that same inspect stability with qualitative data. And this is really important to Glaser and Strauss because at the time they are doing this, there was a resurgence or a deep interest in quantitative approaches in sociology. And so as much as they were arguing or trying to create more ways to talk about what's happening in death and dying, they are also having this subtext tulle argument about the argument about qualitative work. So they described the steps they had taken like this. If you have done research synthesis, you know what the steps are. You make the situation available to be viewed, you describe it clearly. We might take notes on post-its or we might take full transcripts. You make comparisons between parts of your Egyptians until parents become apparent. You describe those patterns clearly and then you say how those patterns describe and explain the world. So when you do this, this is part of what we learned to do the jobs that we have. This was in the 1960s, a radical idea. They taught this by going to small seminars. You could only learn this by hanging out with Glaser and Strauss for a while. And it was usually unusual to do qualitative work this way. It was much more usual to start with an existing theory that had been developed by someone having some introspective insight and then you would use that theory as the lens to view the situation you were examining. So that's more like testing a theory. And what Glaser and Strauss did, they said you can delve a theory just by looking at what's happening. So by grounding their theory building, their analysis in what they saw and what they heard, that was something really new. So that's all well and good. But why does it work? That's a really weird thing to think about. We do it all the time. Clients get value from what we do. We have made the decision to create impact in the world. That piece of research created value for clients or benefits for their customers or users. But it works ultimately because ground theory makes some very careful assumptions about what it means to know
10. 10 things. And ground theory says that you can objectively observe the world and what happens in it. That in these observations there are patterns that you can look at and understand. That the people who are doing the things that you are observing also understand what they're doing and able to describe it to you in some degree of clarity. And that as your understanding grows, what happens is it changes as more information becomes available and that you're change of understanding is also a resource for creating a deeper understanding. So all that can seem very abstract but what's actually happening is it's a form of reasoning which is a fancy way of saying it's a form of thinking. So the kind of thinking you do when you do grounded theory or do synthesis, the way we talk about it, is called induction and induction is one of several kinds of classical reasoning that the Greeks and other people were talking about thousands of years ago. So here's a framework to understand these types of reasoning. When we reason or think about things, we think about elements that are in the world. There might be people or things or both. These elements are arranged in patterns of relationships and when we look at those elements and those patterns of relationships we can see the outcome that gets produced. So induction is about discovering these patterns. And when we do induction, the outcome is we want to be able to make clear statements about likely outcomes. So in grounded theory, because we can observe the world and understand it, in this model here we say well, that's the observed outcome, we can see the elements that are happening and so the thing that is missing, it's like our job, is to describe how those things relate together. And this is also why, if you ever said to clients something about we haven't hit saturation yet, it's because an inductive approach needs you to keep going to increase the likelihood of the pattern of relationships you're seeing being a good and consistent description of the world. So when we talk about saturation, what we're saying is we're not confident yet that our description of the world, our description of this pattern of relationships is strong enough to
11. 11 be a good description of this likely outcome that we're looking at. There is another kind of classical reasoning that is often spoken about is deduction. It's about cause and effect. It's kind of boring because it's about certainty. Deduction starts with the what and the how. And deduction is the process you go through that says these elements, they are arranged in this way, then you always get this outcome. So deduction is about certainty. And this is why people like deduction. It's also why people like quantitative approaches because they feel more certain. So after Glaser and Strauss wrote about grounded theory, they had an intellectual falling out about what it means to create theory. As design researchers you don't need to care. But Glaser, whose equation called it grounded classic theory, said theory is out there in the world and it emerges as you do analysis. Strauss, in his collaborated Juliette Corbin, their theory deduction about how theory gets made. That is a fine distinction to make but Strauss and Corbin say that finding theory is work, you have to put yourself into it. Whereas Glaser is more on the end of it's there waiting for you to find it. And there's a few other versions, in academia people talk about it, Cathy Shamaz and you don't need an opinion about them unless you are doing a degree and even then only if your supervisors care. So most of the time most researchers are doing something had in between versions of grounded theory. If you come from a psychological background than a sociological background you probably what you call what you are doing fanatical analysis which gets rid of some of the baggage of pure grounded theory. So in that reading of it there, is another kind of reasoning that you can talk about and it's called abduction. In the model, abduction says that we have this desired outcome and we know the pattern of relationships that are required to keep the world working. But there's doubts in that pattern that can only be filled by some new element. And so abduction is about figuring out what the elements are. Explaining abduction formally or semi formally like this is kind of weird but it's actually the type of reasoning that most of us use most of the time.
12. 12 If you want to see a movie with your friends, getting that agreement about which movie to see is ab deductive reasoning in action. But ab deductive reasoning is also a creative way to think about your research. So when talking about abduction, one person who comes off is Charles Sanders Pearce with his spectacular beard. He was a philosopher in the late 19th Century. And he had written a lot about induction, abduction. He says you couldn't make up new theories with them, you could only explain existing theories. I find Pearce penetrable. But a few years ago sociologists started writing about problems with induction grounded theory. And they said that a purely inductive ground theory explanation is if limited because you are taking the elements in it for granted. And so like Pearce, they say if you do a purely inductive approach, you're only going to magnify your existing understanding. So here's an example. Suppose you're interested in how come some people find it really easy to use technologies and some people don't. And to investigate this you might look at some extreme examples and so you might choose to look at younger people and older people. If your model explains what you saw is based on the categories of younger people and older people, it's just a trap. It's just a magnification of your current understanding. What you need to do abductively is find a different set of categories that helps you explain the outcomes. One way to think about that, it might not be about older and younger, it might be more about people who can see what's common between technologies and people who can't see what's common between technologies. So you can take that easier and harder out of your description now. It's being able to translate from this artefact to this artefact is where the ease comes about. Easy is an outcome rather than a category. We might still find that older people are less able to do this translation but we're not using that as part of the model any more. We've found new elements to explain what we see. And so according to Timmermans, that is a more creative way to think about
13. 13 things. It doesn't just create a new model, it creates a whole new understanding that you can explore in a variety of ways. There is actually another kind of abduction that people talk about. And you can iterate through this understanding. Instead of having a pattern of relationships that you're looking at and say what elements happen, if you look at both of these things at the same time, and you iterate between elements and understanding and elements and patterns, maybe this pattern explains these elements, maybe this gives us a new pattern to look at, you can create another kind of exploration. Kay Dorst has a book about Creative Exploration. So how do you get better at abductive analysis? It's easy, you just have to read more. And you need to read widely and ideally you need to read about research outcomes. Another word for research outcomes is theory. And so the more theory that you've read, it gives you more different ways to understand the world and the more different ways you've read about to understand the world, the easier it becomes to think about all those different ways to create understanding. And so that will make it easier for you to think in more expansive ways as well. Sam Ladner, who wrote a book, chapter 2 is about applying theory to thinking about corporate ethnographic situations and so I highly recommend that. This is another argument for wide reading and it relates to how we see ourselves. So Timmermans say that we occupy a certain position and that position influences how we interpret the world. There is nothing wrong with occupying the position. You can't not occupy the position. But understanding your position and how it differs to other people's is really important because your position gives you the everyday theories that we explain the world to ourselves as we go. And so really you can't help but see the world through gendered and racialised lenses as you go about your day. And so to minimise the influential lenses, a researcher needs rich theoretical backgrounds to rely on them to get themselves away from their every day categories they can't help but inhabit. So creating new
14. 14 explanations start with describing what we see in terms of existing explanations but then in the abductive sense being able to extend those, to ask questions, not just of the patterns we see but of the elements we see forming into those patterns. And so the more explanations you're familiar with, the more new explanations you will be able to create. Finally, there is another way to get better at more creative abductor explanations and that is to take part in an inquiry. You are doing that now. Talking to other researchers and designers is a great way to grow. In academia people will share drafts of papers, it's harder to do in the commercial world. But if you're in an organisation with more than one researcher, you should share your explanations and half formed explanation and with clients and with designers and the business analyst you are working with as well. There is another way and that is to create a small circle to create ideas. That is hard to set up but if you can sustain it, you will grow as a researcher. Design research is a craft. It's a skilled activity you can only get better at while doing it. It's really hard to teach and learn other than by doing it. And so people who practice any sort of craft necessarily embed themselves in a community of practitioners who all encourage each other in their development. And that's another great way to grow and progress and get better at richer ways of thinking as a researcher. So it's an adventure to get here. Here is a little look back. We do research synthesis in post-it mates because we were told to in the book Contextual Design. They placed that on a bunch of earlier books and some of the interesting once are computers in cognition by Fernando Flores and what they were trying to do was argue the way people know things is different to the way we assume people know things. And they were trying to direct our attention to that gap between what we assume about people and what they really do know and say. And finding out that gap is really deep in our methods as design researchers. Edward Hutchins looking at naval navigation was thinking about collaborative negotiation. When you are
15. 15 standing in front of the mirror board, it's the fact that the research outside your head is extending the power of thinking. First it means that the thinking is taking place collaboratively out in the open. There is power in that for your team and stakeholders. Secondly, it means you can have ideas you wouldn't otherwise be able to because the ideas sit out there. So you can have an idea half formed and leave it and move on to another idea and come back to the first idea later. And, finally, in Glaser and Strauss's grounded theory, the point is you can see what's going on. If we bring ground theory into our collaborative environments, we can find new patterns and new explanations of what we can counter as we do research. So if we look back to where our methods come from and where they are adopted, we can find new arguments about what we're doing now and we ask richer questions about what we're doing now and be more able to completely explain ourselves. So if we understand where we've Dom come from, we will be able to understand where we can go. We can grow our craft and get better. Deep in our methods is this underlying theory about what it means to know things. That's called an epistemology, a theory of knowledge. More than that, design research, doing the design research the way we've come to do it is a statement about what it means to be human and how that is different due to technology. And about how those two things can shape each other ultimately. So if you take nothing away, remember there is nothing as practical as a good theory. And so thank you once again. It is time for questions.