Welcome to "AigoraCast", conversations with industry experts on how new technologies are transforming sensory and consumer science!
AigoraCast is available on Apple Podcasts, Stitcher, Google Podcasts, Spotify, Podcast Republic, Pandora, and Amazon Music. Remember to subscribe, and please leave a positive review if you like what you hear!
Michael Hautus is an internationally recognized expert in quantitative and experimental psychology. As a professor at the University of Auckland School of Psychology, he specializes in sensory and perceptual systems research, cognitive modelling, and analyzing detection and discrimination judgments.
With over 100 publications, Professor Hautus has also contributed to research in sensory evaluation, auditory neuroscience, pain research, human memory, and decision science. His highly-cited works include the 3rd edition of Detection Theory: A User's Guide and the specialized software SDT Assistant, which enables broader implementation of advanced techniques of signal detection theory.
Through his extensive publication record, editorial roles, conference presentations, and collaborations, Professor Hautus has established himself as a leader in psychophysics and detection theory. His current research projects continue to disseminate cutting-edge techniques and push boundaries in the global research community.
Email: m.hautus@auckland.ac.nz
Transcript (Semi-automated, forgive typos!)
Danielle: Michael, welcome. It's great to have you on the show.
Michael: Thank you, Danielle. A wonderful introduction. I'd like to say that I cannot think of anyone to have a bit of conversation with on this topic either.
Danielle: Thank you.
Michael: Thank you for inviting me along.
Danielle: You're welcome. Well, some of our listeners might be new to the field. So one question before we start with talking a bit more in-depth. Can you briefly explain what psychophysics is and how this applies to our everyday experiences?
Michael: Certainly, psychophysics, if we look at the history of psychology, is the original version of psychology, if we go back to Gustav Fechner in the 1880s, who's often seen as the father of psychology. His main interest, apart from many areas that we can call science at the moment was in the relationship between the physical world and human perception. Essentially, that's what psychophysics is. Psychophysics is the relationship between physics, the measurement of the physical world, and our perceptions, our psychology, and our relationship to that world. In that regard, it covers pretty much everything that's human. Everything comes in via those processes. But more commonly, we talk about psychophysics in directly measuring physical stimuli and seeing what their perceptual counterparts are and trying to unravel the relationships between different dimensions of physical stimuli and those perceptions and the emotions and things that right on top of it.
Danielle: Okay. Yeah. Thank you for that explanation. My second question is also a little bit more about the background. We're always curious to hear about the parts that our guests take in their career. Could you share with us a bit more of the story about your professional journey and how this brought you to your current position?
Michael: Sure. My professional journey started with me beginning at the University of Auckland in 1984. One of the things I did back then was I thought, I'm going to change what I've done from what I've done previously as a school, which was all sciences. I did a year of art subjects, and I felt some of them. I won't say I felt miserable, but they were just not my cup of tea. I re-engaged with science and in particular in psychology and undertook my undergraduate degree, where I was more interested in the clinical aspects of psychology at that stage, and found that that also wasn't long-term for me. And so switched into looking back at the things I've done in the past, and a lot of that was topics like chemistry and physics. And one of the things I've really loved was computer science. And so in psychology, I moved into the more scientific end of psychology, I guess you call it the more quantitative end, and where you could apply things like physics, and you love computer programming, and things like that. And so that end was psychophysics for me. And so ever since my master's degree, I've been working within that area. I've always been at the University of Auckland, so all of my degrees were there, and my career has been there. And it has been a continuous evolution of different topics within psychophysics and sensory science, more generally, that I've followed. It's taken me on a lot of different tangents, just exploring things I was interested in off to the side, things like pain, perception, music perception, and all sorts of other different areas. Essentially, that was my path to doing the things that I love to do, finding what they were, and exploring them.
Danielle: Sounds great. Yeah, it's really such a broad range of topics that you worked on indeed. All the way, I mean, we worked, of course, obviously a lot together on sensory, but it's so much broader than this, the application of psychophysics.
Michael: Yeah, it goes back to that original thing I said, what psychophysics is. It's pretty much the relationship between anything in the world out there that can be measured and your perception of that. And so the common core elements of psychophysical theory, if you want to call it that, can be applied in so many different ways to the world of sensory information. But a lot of those laws can also be applied to the world of information that comes out of sensory information. By that, I mean, you can have computers making judgments about the world and things like that, or simple devices like smoke alarms that give you information that they extract from the world, and then you perceive that there's a relationship, a psychophysical relationship between you and the smoke alarm or any alarm system. That's another layer on top of that. So it's not just sensory stuff. It goes a little bit further.
Danielle: Yeah. Building on that with your experience in psychophysics, have you seen also new technologies, and how they transform the research in sensory and consumer science?
Michael: I guess the first new technology that I saw was computers if I go back far enough, and how that revolutionized the way things were done in the boroughs. And of course, now, everyone takes computers for granted. But when I was growing up, you had the first portable computers that were made for the consumer market that you could get. I remember writing a PhD on one of the first laptops you could ever get that had two floppy disks that were inserted into it. They had no hard drive or anything like that. And that technology has always been fascinating. And of course, that technology has evolved over time to get more sophisticated. The things that one used to do on computers in the 1990s, where they were extremely slow, I remember running programs to data that took two weeks to run, whereas now that will run in 30 minutes quite easily or less. That transformation is amazing. Then there's other specialized technologies that have come into play over the years, especially brain imaging type technologies, EEG, and MRI. I've done quite a bit of EEG over the years. I've always been going to go into the MRI, and there's a story in that later maybe. The most recent technology is FNAS, so that's nearer, with respect to Scotland, which is relatively recent. It's actually very amenable to doing work in sensory because you're not constrained to not move, not use your muscles, and things like that. It gives you this access to ecologically valid, realistic human behavior and taking measurements of brain activity.
Danielle: Yeah, that was always the limit that you couldn't even chew or have people evaluate a product. All those limitations.
Michael: Yeah, well, for EEG, that's a major problem because those are huge muscles that are being used to generate a lot of electrical bills. And fMRI, it's also a major issue. We did run a project several years back. We had a colleague of mine, Nicolas Cant, came up with some special scanning, if it's fMRI, which enabled us to actually do some recent work in that field, but it's still extremely difficult.
Danielle: Yeah. Well, on a different topic, in your work, you developed and refined, and we worked also together on that, sensory discrimination test. How do these tests contribute to your understanding of sensory perception? And how do you see the role of technology to enhance these tests?
Michael: It's an interesting question. My perspective on sensory discrimination tests or broadly methodology. So any approach to collecting data and analyzing has always been from a perspective of wanting to create better ways of measuring a sensoryl phenomenon and quantifying those. I've always come at it really from that perspective. I want to see measurements that are valid and they have small variations in them, and they can be used to tease apart things in the real world that we want to know about. The other side of that is that these tests are extremely important to be used out in the world, and it's so slow for people to adopt good methods. It has always amazed me, and I'm not the only person that says that, is that people tend to keep doing what they've always done and think that what they're doing is great. Sometimes it is, and sometimes there are a bit of things that people could be doing. And this is very true of methodology and how people go about collecting sensory data and how they go about analyzing that data and using that information. From what I can tell, the whole process is very inefficient, and people are spending a huge sum of money, and a huge amount of resources collecting information about different types of sensory stuff, but sometimes it's companies looking at their own products and looking for ways to actually make more money out of this. But at the same time, they're spending a lot of money and wasting a lot of it because they're not getting as much efficiency as they could get out of what they're doing. On the practical side, one of my goals is always to try and get people to use the right methods in the right scenarios. And I think that's one of my goals for the next several years, is to try to produce some more literature, some more resources to help people to do that, to know that what they're actually doing may not be the best way to do something. Move away from what has been done in the past, which, as I say, some of it's great, but there's a lot of new things, a lot of new ways to do things. That's how the world will change if people start using more efficient ways of collecting information and of using it.
Danielle: Yeah, that is what I sometimes feel that the field of sensory in the sense, on the one hand, is very modern and looking into new things. But for example, in looking at how contacts affect it, whether we should look at the user experience instead of just one of tasting in a lab. In that sense, people think about how to improve, how to make it more ecologically valid. But if we talk about test methods, one of my biggest, well, I wouldn't say frustrations, but almost frustration, is that when we talk about signal detection, for example, people immediately start saying, "Oh, but that is very complex." While I think that the traditional statistics, at least for a person like me, is at least equally complex with all those formulas. I still don't get what is the big hurdle in this field of signal detection, because it's also, for me, it all is so logical. Do you have any ideas around this?
Michael: It's an interesting question because, and I see that everywhere as well, people are quite happy to use standard statistics on things. Maybe that's because they're actually taught standard statistics. When we're getting the qualification, so they feel comfortable with it. Whereas if you look at signal detection theory, signal detection theory is almost standard statistics, just done slightly differently. But people are not taught it. Even in psychology now, you can go through psychology textbooks that use stage one, two, and three. You'd be hard-pressed to find anything in there anymore about signal detection theory. It's just too complicated by authors to actually put into it, even though you have specialists people writing textbooks for psychology that are on research methods and analysis. But there's a nice overlap with statistics. So with statisticians and statistics departments in universities. And so you've got that connectivity going on. It means it's easier to get people to understand and to, well, not so much to understand because both would be signal detection through the statistics are relatively similar to understand. But you've got this pathway for research-based statistics that engages people because they have to engage with it as expected. One of the determinants of signal detection theory or modern signal detection theory, one of the strongest influences on it was actually Neyman-Pearson's approach to statistics. It's just a reworking of that approach to statistics. It's interesting that people do resist it, and I find that all the time. It's a thing that runs through in psychology, at least where, I mean, it was the start of psychology. Beginning, psychology emerged out of psychophysics. And yet, psychophysics is now I was left behind to some extent. And signal detection theory is deemed too complicated to expose psychology students to. I ignore that, and I teach them to do that. I just go ahead, and they get it drummed into them, and then they enjoy it. They think, "Oh, yeah, it's a great way to think about things." Then when you think about how it's done in other places where you ignore things like just the basic sensitivity and response bias and separating those two things, they start to wonder, "What are these people doing?"
Danielle: Talking a little bit more about this detection theory. So your book, I think, is probably the most important book in this field.
Michael: That's very kind. I don't know if it counts as the most important book in the field. That sounds nice.
Danielle: At least, for sensory, I would say for sensory consumer science, if people want to know about signal detection and how it applies, this book is a good start. It's not easy, but no, it will be, I think it's really a very important book. But can you share about how detection theory has evolved with technological advancements and its impacts on food and beverage and cosmetics?
Michael: It's a tricky question because detection theory is just a way of doing things with a bunch of models and stuff in the background. I guess the way that it has advanced in recent years, well, all the theories keep on advancing, and they get into more detail on things that we're not really open to. I'm concerned about discussing that stuff at the moment. But where it does seem to be getting easier to use is in the software that's made available for people to engage with signal detection theory. There are now a whole bunch of packages in R that do various sorts of analysis. And there's a software that I put out called the SBTA system, which does a lot of different analyses as well. I think that the tricky part is people don't know how to use this software and what they should do with it. But it's out there, and quite often it doesn't come with a standardized framework that works consistently across different signal protection approaches. And so that's something I've started to focus on a bit more, and I'm writing a paper at the moment that is designed to just take a basic look at two tasks, the A-not-A task and the dual two-alternative forced-choice (2AFC) task, and look at the model that's involved in those and come up with measures that are consistent across the two tasks and of course, d-prime is an obvious one that's consistent because that's the standard sensitivity measure for the signal detection theory. But then what about bias measures? And so the idea is that you should be able to extract a similar response bias measure and use it across the two different tasks. And that should be then extendable out into other tasks as well, like saying different or the reminder tasks or any other task that's based on psychophysics that you would use to collect data. That's technology in the sense that it's refining the ability is converging the computational strengths of signal detection theory, making it more accessible, which, speaking back to what we were talking about before, people not wanting to use the stuff, will make it easier for them to engage. I think the next step is to make it clear, our way to engage. Not just throw software people and books, but have some instructions of some kind. One of the things I like to do is put together some webinars, talk training sessions that use startup, this SBT assistant, probably, and show people, "Here's your data. Now, this is what you do with it, and this is what you get out of it." And actually plug it into the software and show it being done. I think that will make, for those interested, it would make it easier for them to get the methods.
Danielle: Definitely. I think it's that comparison of methods and indeed not just in d-prime, but also in other parameters. If we think about it in the future where there is more emphasis on reusing data or knowledge management, we will have to go to that in order to connect the results of different studies and everything.
Michael: Yeah, that's the one thing that, I would say is the one, that's one of the big advantages of using signal detection theories. It gives you a common metric that you can use across so many different, not just sensory, tasks, you can use them in anything we have diagnostics. And the sensory task is a diagnostic task, essentially. And so you could use these things in looking at how well algorithms can provide answers to questions that you have, because they'll either provide correct answers or incorrect ones. So you could come up with the basics, its hypothesis, also known as correct projections for what AI tells you. And it can be used anywhere. And you can come up with a common metric that you can use to look at the relationships of the different types of research to go across different products, it's all across different companies all over the world. It's a nice big database that people were willing to share some knowledge. Things would move so much faster. Companies would make much more money. And moving things to be more healthy would happen a lot quicker as well. There's so many benefits that can come out of having a shared index of how well people can do something, how well they can tell things apart.
Danielle: Yeah, I think so. Indeed, if that would be shared across companies, and it's not really that you then give away confidential things. I mean, not necessarily. It is just everyone is looking for the same direction, more healthy, more flexibility in ingredients. Yeah, it would be great if that would be possible. But I would also like to talk a bit more about other research because now we focus really on the core sensory. You also do research on binaural tones and their role in stress reduction. I'm really intrigued by that. Can you tell a little bit more for our listeners what you are doing and also what the impact you expect that this could have on wellness and health care?
Michael: Sure. Binaural tones, or binaural beats, as they're probably more commonly known to the public, are essentially just at the basic level, two sounds that are made of a single sine wave. So they're the simplest sounds that can be produced. And they are slightly different in frequency. And you take one of these tones and you put it into someone's left ear, and you take the other tone and you put it into someone's right ear. So you've got two sources of information coming in. When the information from both ears converges within the brain, you end up hearing an illusion. It's a beating illusion. The amount of times that a beat happens per second depends on the difference in the frequency of the times. You can set up these beats to perhaps go at 10 cycles per second, which would actually be close to the ultra-attivity of the brain. You target these tones to specific major ranges of commonly partitioned brain activity. If you like to put it that way. And so the idea is, or the common idea that's publicized all over the world in so many different places, well, this will entrain your brain, and your brain will follow those frequencies. And so if you target alpha bands, then your alpha will become more synchronized, and that has natural health benefits.
Danielle: And is it true?
Michael: Well, the research we've done so far seems to suggest there is some truth in that. What we've done is essentially put people through a minor stress test. So you give them a task that's a little bit difficult and a little bit stressful to do. We often use a thing called the Paced Auditory Serial Addition Task (PASAT), which is the paced auditory serial condition task. You essentially sit in front of the screen, this is a visual version, and you're given individual numbers, and you have to add each successive number to the total that was previously. You might have someone standing behind you telling you to do it right, get it right. So you put the person under a bit of stress.
Danielle: A little bit more stress.
Michael: Yes. Without getting too carried away. And then you have a period after that where the person just relaxes and you measure, let's pick one measure, let's say the galvanic skin response and how that changes over time. And what you find is it changes over time, the skin conductance improves, and it gives you an index of how quickly a person is relaxing. So this opens up opportunity of actually doing other things in that silent period where you're measuring the GSR. You could also present binaural beats or music or anything else you might happen to think might have an influence on the recovery period. And so what we did was we presented binaural beats of different frequencies to participants during these rest periods. And we found that low-frequency and high-frequency binaural beats actually did improve the speed of recovery. We didn't find that alpha ones, that there wasn't a significant difference there. So there is an effect going on, but it's possibly not exactly what people are saying.
Danielle: That's how they thought.
Michael: Yeah. So there's a lot more work that can be done in that area. It's quite fascinating because, of course, if you can find and show a functional relationship between listening to tones or certain music and recovery from stress, then you have evidence, I guess, that this has efficacy and it actually works and isn't just a bunch of hype that people spin up and advertise it. We're looking to do some more research in that area. We're also looking to do... We started some research looking at different genres of music and how listening to them can influence the recovery period, and that's ongoing at the moment. Of course, going forward, this could be something that people could use personally as a way of relaxing, and being shown to relax people. Or it could be used in recovery wards, hospitals, rest homes, and all sorts of places where it's helping them to settle and relax.
Danielle: Yeah, in waiting rooms or something. The doctor if you're stressed with the dentist.
Michael: Yeah, absolutely. It's an exciting area, and we are certainly going to pursue that somewhere.
Danielle: Interesting. I see that we already need to... The time flies, Michael. I think I need to invite you sometime in the future for the second half. We have to wrap up a bit.
Michael: Sure.
Danielle: Oh, many good questions that we need to postpone. But yeah, one thing that we always like to ask our guests as well is for young scientists and researchers that are interested in psychophysics and sensory science, what advice would you offer? Especially maybe also related to leveraging technology, but what advice do you have for young researchers?
Michael: It's a tricky one. So this is my personal advice. It's slightly different from what most people might say, which is work hard, and do what people tell you to do. Whereas my advice is to follow what interests you and do not listen too much to what people are going to say because everyone just keeps doing the same thing. Find something different. Find something more. Find something better, engage with it, pursue it, and you'll make a difference. I think that's, here we go, that's my piece of advice.
Danielle: Yeah. Well, that's very valuable. Thank you for that. Then my final question as a wrap up, if our listeners would like to reach out and get in contact with you, what is the best way to get in touch? Should they connect on LinkedIn or send an email?
Michael: The easiest way would be just to send an email to me. I do occasionally look on LinkedIn, but not as often as I probably should. And whereas by email or going via my website, https://hautus.org. There is a form there that can be filled in and reach me through that as well. But everyone is welcome to ask me questions. Be in touch. But those are two methods.
Danielle: Okay, good. Well, when we publish the show, we will then send his details, put his details there so that people can contact you. Okay, well, Michael, thank you so much for this conversation. I really enjoyed it. I hope you come back sometime in the future for the second half of the questions.
Michael: That's been fun. And yes, these are things that I can talk about for a very long time if one is not careful. And so, yes, more questions in the future would be great. Thank you for having me on the program.
Danielle: You're welcome. Thank you.
I hope you enjoyed this conversation. If you did, please help us grow our audience by telling your friend about AigoraCast and leaving us a positive review on iTunes. Thanks.
That's it for now. If you'd like to receive email updates from Aigora, including weekly video recaps of our blog activity, click on the button below to join our email list. Thanks for stopping by!
תגובות