#43. Subjective Fear and The Experience of Presence - Dr. Christopher Maymon

Dr. Chris Maymon, of Victoria University of Wellington, talks to us about his research on the experience of being Present. Him and his team use virtual reality to test the causal role of subjective and physiological components of fear in generating presence. Dr. Maymon goes through the background of understanding being present as a concept and how it is measured, the use of virtual reality in his experiments and some of the interesting findings including how your subjective fear and not your heart rate and other physiological measures impact our experience of presence.

Read the paper here: https://www.researchgate.net/publication/379928195_The_presence_of_fear_How_subjective_fear_not_physiological_changes_shapes_the_experience_of_presence

Chapters

• 0:00 Introduction to the paper
• 6:00 How do you study presence?
• 19:18 Why use virtual reality to study presence?
• 35:05 Top down and bottom up processing
• 44:10 Mediating relationships in Presence
• 50:36 Theories for why physiological changes don't explain presence
• 1:01:12 Explanations and future directions

Transcript

0:04

Awesome. All right. Shall we get started? Fire away. Welcome, everybody, to the Smooth Wind Society. I am Sahir. This is another episode where we're talking about a particular research paper. And today we're going to be talking about a paper titled The Presence of Fear, How Subjective Fear, not Physiological Changes, Shapes the Experience of Presence. Now, this is written by Dr. Chris Mayman and colleagues from Victoria University of Wellington. Some of you might remember that Dr. Mayman came on before and spoke about virtual reality a few episodes ago. So he spoke about how virtual reality is used in neuroscience research and cognition research. So glad to have him back on now talking about his paper as opposed to... more general idea of virtual reality and research as he did last time. So welcome Chris. Thank you, thank you, Kia ora. And as always, you guys know the format, we get someone on who has no real experience in the type of research. So we've got Vroyden back on. If those, some of you might remember Vroyden was on a long time ago when we were speaking with Josh Faulkner about concussions. Since then, Vorden has done a lot. So now he is doing his PhD himself in computer science, looking at how we can generate and understand emotional movement. So I guess there might be some relation to Chris's work, but not a lot. I did ask Vorden if he knew much about virtual reality and research, and he said he did not, which makes it perfect. But it also means that he can ask some questions from a computer science side of things as well, which would be really interesting. His interests range from technology and AI to consciousness and neuroscience and to the ways we can integrate artificial intelligence into our daily lives and skills. He's also a pretty decent basketball player and he says that he enjoys yoga but he does not look like a flexible human being whatsoever. He says why I enjoy it. But awesome, Royden, welcome. Thank you for coming on as well. Thank you very much. Great guys. So Chris, let's start with a little background on the paper. So could you tell us what the title means? And yeah, how you got into this? Sure, yeah. So this is one of those titles that really tells you about the all the results that are the most important to know about, to understand the conclusion of the paper. So the first part, the presence of fear. It's just alluding to this puzzling relationship that has been observed in past studies between feeling afraid and feeling the sense that you are present. And not really knowing whether this is a relationship that is driven in one particular direction or the other, or if it's bi-directional and there's some sort of feedback loop going on. So that's what the first part is about. And then the second part, how subjective fear, not physiological changes, shapes the experience of presence. This is about the findings in our study. So we found that how afraid people say that they are, which is the subjective component of feeling afraid, that aspect where if I just ask you how afraid you are and you tell me on some Likert scale, that is your subjective emotional. level. It's that component of feeling afraid and not the changes that happen physiologically that are also associated with being afraid. Your heart rate increasing, your hands becoming sweaty. These are all part of the fear responses, but it is not those changes that are influencing how present you feel. It is the changes that are happening subjectively to you. And the reason that this is so, you know, the reason that this is a new finding is because actually theories about where that feeling of presence comes from. And when I say presence, I'm talking about this kind of sense of being there, you know, the sense of being in the moment, in your body. Yeah, in reality. Is there are theories that say that this is really driven by feeling changes in your body, by interoception is what that is called. So there are these signals that are coming up from your viscera, your body, your heart, your lungs, everywhere else, you know. And you feel these changes on some level and they're part of all of your emotional responses and theories about presence say that that's a direct contributor to... to what it is to feel present. And our study does not show that. It shows that instead it's really purely a subjective to subjective relationship. But it is not one that sort of crosses into physiological changes that are predicting presence. So actually, our paper is perhaps more about a relationship we do not find, then it is about the relationship we do find in a lot of ways. Should I ask some really basic stuff which I'm trying to wrap my head around because you said these are the key findings that you found. So by presence, how are you measuring presence? Because you said it is once feeling of being there or being in a situation. So how are you measuring this? Yeah, this is a super, super important question actually. Um, and I'm glad that you're starting with this one, um, because measuring presence is something that we are very, uh, limited, uh, in our ability to do it. We really only have at our disposal questionnaires and of course we could, we could collect ratings. Um, so, you know, one to 10 ratings when we just define presence for people. And. So it would be wonderful if we had some sort of EEG measure or a peripheral physiological measure, something in your heart rate variability that would be kind of a correlate of presence, but no one has so far been able to identify one. And we also were not able to identify one in our paper. That's not to say that there isn't one. It's just if there is one, it has yet to be identified. Which, which makes this really tricky. It kind of motivates questions about, well, how do you know that it's really. That it exists, right? Like, how do we know this isn't just something that we kind of, uh, that it's more of an artifact of the question. You know, if I just ask you like, how. Red do you feel today or something like that? Like you might give me a one to 10 rating, but it wouldn't like mean anything. Right? Like some people wonder if presence. you know, it's sort of like that. And yeah, there's, I think there's merit to those points. The only thing that is perhaps, the only thing that perhaps suggests that presence really is a real phenomenon is one, there are these relationships between presence and certain emotional changes that are robust and replicable. So we're not the first to find that subjective fear. and presence are positively related. And also there are disorders, dissociative disorders that are characterized by a reduction in presence. So depersonalization and derealization disorders are disorders characterized by feeling like you are either not in your reality or that reality is somehow not the real reality or that you are not the one controlling your body. You're not actually embodied that you don't have the kind of agency that we all, that most of us walk around and, um, experience day to day. Um, so yeah, in terms of getting at, you know, what is presence, how do you measure presence? Yeah, this is where this is, this is a question that I think gets at the, the infancy of this particular scientific question. Um, there are, there is a need to try and understand how to measure it better. Um, in our study, the primary measure is to ask them how present they feel on a scale of one to 10 at different time points. Um, I could stop there, but this also lends itself to the question of why we chose to do this in VR. Um, so not, not to, not to suggest a question because I know that's your job, but at some point I should probably, uh, clarify that cause that also plays into this measurement question. Yeah. I was wondering how does presence relate to the idea of like being in the zone or like a flow state and that kind of thing as well. Yeah, I think that they are related, but there's a lot of different... Either cognitive functions or, um, sub or psychological phenomena that people ask, you know, is presence related to that? For example, yes. A flow state is one of them. Attention is another one. Like how much is, is presence just being very, very attentive? Um, and there's, there's really, there's, there's not, there's not a lot. There perhaps is no good research that can really clarify where the divide between these two things, um, actually. lies. I do suspect that when you are in a flow state that you are highly present, it's about a feeling of being in a moment. And people have also asked like, well, does this have effects on the passage of time? And I suspect that it does, but I don't know if it's presence that's doing that direct... Sorry, I don't mean the objective passage of time. I mean the subjective passage of time. And so like, first of all, I should say with questions like that, what we're probably going to encounter a lot in this episode is that I can really only speculate at best. But I, the thing that I speculate perhaps more confidently is that a lot of these effects perhaps go, are mediated by emotional change. So I think that there's probably a relationship between presence and whatever your emotion that is kind of driving that change in presence and that through that emotional change, there might be changes in say time perception or flow state or attention. And so I don't, so I'm less confident that there would be a direct relationship between presence and and flow state and attention and things like that. But again, purely speculative. Just to derail the topic, do you guys both watch anime by any chance? I watch very mainstream anime. I've seen a lot of it, but I haven't watched the last day of the end of it. Because the moment Royden said Flowstate and Zone, all of a sudden my head went to this anime called Kuroko's Basketball. where all of these characters basically get into a zone and their focus just increases and they just start doing magical things like scoring points from beyond half court and stuff. Somewhere. Yeah, that's the first thing which came to mind and the other thing which came to mind is kind of like when you kind of drone off in class or in lecture, something when someone's giving a presentation you're drawn off and then you realize you've not been paying attention and then all of a sudden you try to hyper focus and pay attention. Is that also where kind of that idea of presence comes in that oh I had kind of like I was not aware of what's going on and then like you touched on attention being like then you try to focus and be present in the situation. Yes, yes. So one of the things about presence that fascinates me the most is the two realities. So if you read a really good book or something and you get really drawn in, you know, you might not notice that hours have passed. You might not notice that, um, that you're actually like emotionally, you're having emotional responses in line with like the protagonist of the story and not in line with you sitting like on your bed in the middle of the night, right? Like And I find this to be one of the most fascinating things about this sense of presence. And, and really it's the ability for our minds to become absorbed in, um, realities that we create. You know, in our, in our minds, um, and that are, are physiological and subject our, our bodies kind of align. with this reality that we, that we generate in our minds, but it's not necessarily the reality that is informed by all of our sensory modalities, the way that we do when we're walking around, you know, not daydreaming. Right. And so this is, this is the thing that I'm. Yeah. Probably the most interested in, but I've yet to really come up with a really good way to study that aspect of it. Um, yeah, cause I mean, this has happened. I think this is. what drives like the video game industry is the ability to feel, um, so compelled by a medium, um, that you kind of get absorbed into it. I mean, otherwise, why would people spend so much time, you know, playing them, right. And so much money on them. Um, and yeah, and yet we really don't understand what, what is happening. I mean, it's, it's more than just attention. If it was just attention, then anything could do this, you know? Like people like, yeah, like, like it's, there are certain things that lend themselves to being, to being immersed into more readily and for longer periods of time, um, than others. I mean, that might be a interesting set of studies. They're just a bunch of things on time perception. video games and just anything else. Also like their response on VR things. And then just looking at time perception is a marginally more objective measure. Yeah, there's actually a study that I'm part of. It's being run in Vancouver at the University of British Columbia right now. And it involves a simulation that we created in our lab, a VR simulation that is meant to induce awe, the emotion awe, A-W-E, in people by showing them this kind of vast, fantastical space or underwater world. There's a few conditions. And the idea is that we're trying to measure the passage of, changes in the perception of the passage of time relative to a control condition. And I'm eagerly anticipating an email telling me what the findings were. because they're at the last couple of participants being run right now. Yeah, so we're trying to figure out which emotions might be driving changes, or if there are emotions that are driving changes in time perception. But then after that, I'm really interested in looking at that relationship that I was describing to you before, between presence and emotions and time perception, and which one might be more directly or indirectly kind of contributing to this effect. Could you have a time perception element in this? Because you said presence of fear. So would you put potentially if you're looking at someone's presence in a situation where you're inducing fear, could a time perception be added into your study? I haven't even gotten into your study yet. No, yeah, it could have. I mean, one simple way is just to ask people at the end of the. after leaving VR because they're only in virtual reality for about three or four minutes, depending on how long it takes to do the second part where they have to walk across a piece of wood. You could ask people how long they felt that they were in virtual reality for, and we would know exactly how long they were in for. So yeah, we could have asked that. I keep coming across things where I go, I really wish that I had included that, that measure in the study. I could have looked at this and I could have looked at that. But, you know, one of the things that we never talk about as scientists, because I think it's kind of, uh, it's more, it kind of reveals some things that perhaps we're all insecure about is that we. Miss opportunities every time we do a study. And this is, I think inevitable. I don't think that it can be helped. And every time I do a study, I'm just. I know I'm going to go to a conference and someone's going to ask me, why wouldn't you just include that as a measure? And I'll go because I didn't have the thought to do it. And I wish I did. This is one of them. Yeah, but hopefully I'll run this study again, because there's actually so many other ways that I could have tried to unpack this relationship differently. And now that I've seen it this way, I intend to. Okay, so. Instead of asking you of things you could have done, let's talk about the things which you have done. I think that's a nicer way to go about things. So just looking at the abstract and all, there are multiple studies in this paper. It's not just one study. There's a few different parts to it. So where did you want to start? As you mentioned, you wanted us to kind of ask you why VR to begin with. So maybe, could I pretend? Yeah. to ask that question and you can tell us why you use VR to look at this? Yeah, so actually I have a simple demonstration about why. So here, on a scale of 1 to 10, how present do you feel right now? In this conversation I'd say about an 8 maybe? And eight. Yeah. And Royden, if I was to ask you right now, honestly, like scale of one to 10, how present do you feel right now? What do you think? I'm like a six or seven. It's just much earlier than I'm normally. Oh, okay. Well, you know, you guys just did not demonstrate what normally gets demonstrated when I asked this question. Because usually when I asked this question to like students, because sometimes I have a visit, there's one class here that I teach where I talk about presence a little bit. And usually when I asked this, I get nines and tens. And, you know, this is what I've come to expect as outside of VR. If you ask somebody out present, they are they're going to give you a really high number. The problem with that is that you don't have as much variance with which to look at changes specifically in the upward direction. And when you put people in VR, a convenient thing happens where, today or the last couple of years, because we ran this study back in 2020, people will not give you numbers like eight, nine, and 10. They'll give you numbers like four to six. And this is really good, because then I can look at factors that might increase or decrease it. And so really, the reason why I put it in VR is ultimately just to have a more meaningful a better variance in my presence ratings with which to look at change. And so what they, can I, can I say what they do in VR? Cause I feel like I probably should like give people an idea of what the, yeah. So if you're a participant in my study, um, and this will be true of the first study as well as one of the two conditions in the second study, you go into the virtual reality and you find yourself on, uh, the sidewalk of a busy city street. And it's a, you know, there are these high rise buildings all around and it's not a city that you'd recognize. It's totally fictitious. And there's a high rise building behind you and you can hear the experimenter's voice there outside of VR, obviously, and they're just asking you. Okay. Take a few steps forward and you take a few steps forward and you get used to seeing how, Oh, I'm in this headset, but you know, the world moves around me. as I would expect it to, you know, that you need to have this period of time in a VR study where people get acclimated to just moving around in a world when everything's mediated by a headset. And then we ask them for a set of one to 10 ratings about presence, but also about different emotions that they might be feeling. So relaxation, fear, excitement, sadness, anger. things that we are interested in and things that we also don't expect to change. Um, that might have the same valence as the thing that we expect to change. So valence being how positive or negative is the emotion, or that might have the same arousal level. So like, you know, high arousal, low arousal. Um, and so we're going to capture these ratings about five or six times during the study. Um, I should also mention that the entire time they're in VR, they are wearing this wireless. vest under their clothing, which measures their heart rate and their skin conductance and their breathing rate gets a sense of their physiological condition. So they walk around in VR for a couple of seconds. They see this city, they provide some ratings. We asked them to turn around. They said there's a building that was behind them. That building is a little strange because instead of having a normal front door, it has an open elevator. And so they walk into the elevator. They turn around. big red button on the inside of the elevator. They provide some more emotion ratings here. Again, nothing should have changed. This should be just like the part of the neutral environment. And then they can use their hand controller, they're holding in their right hand to like push the red button, red button, the doors close, the elevator rises, rises like 60 to 80 stories, roughly. It looks like it's about 60 to 80 stories. It's high enough up that it's a lethal drop. And when the elevator door is open, you see a wooden plank in front of you. And we have placed a wooden plank on the floor of the laboratory in the exact position so that it would match the plank that you see in virtual reality. And we've also asked participants before they go in VR to take their shoes off so they can feel this plank beneath their feet and they can actually get a sense for where its edges are and stuff. And we ask them for ratings when they're in the elevator, seeing what they are now about to do. And then we ask them to step up onto the plank with both feet. Just at the start of that plank, we ask for another set of ratings. And then we ask them in their own time to please try to walk to the end of this plank. And they do so. Well, most of them do so. And then we ask for a final set of ratings. When and if they reach that end of the plank. And. In the first study, they were allowed to try and step off the plank, and that would trigger this kind of falling animation. But to be honest, as far as we're concerned, the study is over as soon as they've provided the last ratings at the end of the plank. And so that's the general procedure. And the only difference in, I mentioned that this procedure describes one of the two conditions in study two. The only difference with the other condition is that instead of the elevator being an elevator that takes you up to the top, it takes you about halfway up and then it comes back down and then it opens up where you were before. But now there is a plank on the sidewalk and you have to do the same thing balance on the plank. But now basically what we've done is we've isolated the height exposure as a variable, as a function of condition. And so it's our It's our control condition. that kind of gives you a sense of like the design procedure of the study. I've done the plank study a couple of times with you now and it is quite scary. I did one with Jordan's running, Jordan Shull's running a, I guess a different edition of it at the moment. And in the past when I've stepped off the plank, it didn't drop me. I think I was doing a slightly different version of the same thing. And I asked her halfway through it, I was like, as I was walking on the plank, I'm terrified of heights. thought I'd be fine, but I was walking along and I said to her, I said, if I step off, will I fall? And I think she says something along the lines of, I wouldn't find out. And I tell you what, that was us terrified from that point onwards. I can only imagine my heart rate spiked or something. But I did make it to the end of the plate. It was, that was so cruel of her though. I did also find it really interesting what you were saying Um, and kind of intentionally decreasing the realism of the study. I mean, obviously I was, I was always surprised by how real VR feels. Obviously, you know, it's not real life, but I'm surprised by the realism of it. Um, but also find it really interesting that you're kind of taking advantage of the fact that it is less real. Um, and I guess I wonder, so my assumption has always been that there's been a a want to make these simulations or these games that people are playing in VR more real. So I wonder if that was an option, you know, Unreal Engine 5 and all of these great graphics, apart from obviously being very difficult to drive, you know, great graphics card. I wonder, you know, is that something that you would opt not to do? So yeah, we did consider all these ways to try and make it more graphically realistic. And you're right. It's a, there are just processing limits that are so yeah, you have to purchase this in terms of computational expenditures. Um, a graphics card, even a really good graphics card will struggle to do the kind of like volumetric ambient lighting that is characteristic of a kind of Unreal Engine five simulation. This is especially the case in VR because, and I people know this, of course, because if you think about VR, you realize, oh, of course, you've got two screens, one for each eye. But they're not the same image. It's not like just plugging in a sec, a second monitor. They're actually offset because if they weren't offset, you wouldn't get the, um, the depth perception. You wouldn't get the illusion of seeing into an environment. You would, you would have things where your interpupillary distance isn't being accounted for. You wouldn't have that stereo. Um, vision. And, uh, so the reason that that's important to remember is that it is actually two different images that are being rendered and they have to be rendered at a very high rate of speed in order to reduce, um, the likelihood to make people nauseous. But also there's a certain refresh rate where if you don't reach that people, like the illusion of even being in a place revert back to a realization that this is actually images presented at a high rate of speed to your eyes. And so the faster they go, and they also have to take into account your movement, right? Because it's, yeah, you're looking this way and it's presenting things as things change in the environment. But as soon as you start to move like this, they have to, they have to do that same calculation, but they also have to take into consideration being able to detect where your body is moving. And then redraw the scene from the perspective of those two kind of cameras that it's creating in the virtual environment. Um, and it has to do that very fast. So if you have, yeah, really advanced lighting that just looks really profoundly gorgeous, it's, it's actually going to like eat into a limited supply of computational resources that you just have a very fine limited amount that you can, that you can play with one way or the other. So optimization in VR is a big part of what I actually end up doing in the lab in terms of like hand off hands on work in the lab, it's not actually just about creating stuff. You can find stuff that people have created and they're more than happy to either, you know, we can, we can buy like assets or we can just download freely available assets. Like that's no problem. It's optimizing it for VR. That is really the. crux of the, the challenging work that I think I do. Um, and in terms of your question about if we deliberately, uh, made the simulation less real, uh, I would put it in a slightly different way. We, we realized that it didn't have to get more realistic for people to feel present and to feel afraid in it. So I made a city, you know, as well as I could back in 2019, um, And it did not look real to me. And it didn't look real to a lot of people, but I did this one little thing. I turned down the light. I just made it look like this was a city at like 9 PM, like middle of summer, you know, that kind of like end of Twilight kind of dark, everything now had shadows. And now I hadn't really changed anything about the quality. I just taken light away. Light. It's not really light, of course, because it's in a computer, but you know what I mean? And. And people suddenly felt like this was a much more realistic city. They just thought it was kind of grittier. And it's, and it's, it's funny. This thing that people will fill in empty gaps in, uh, gaps. But I say gaps, of course, I just mean like, if you don't show them something directly, you kind of hide it behind like darkness or something, people will use their imagination to fill it in. Actually. That's almost that's better than anything that could be like photorealistic around them because. They're generating it themselves. So, you know, it's like offloading that computational expenditure onto the participants own imagination, um, which is free as far as I'm concerned. And so, so yeah, it's, it's more, it really comes down to a budgeting choice ultimately. Um, but yeah, it's just one of those wonderful things where people don't need a photo realism to believe or feel present in a scene. I should say, feel present rather than believe. because I don't know if it's really a matter of belief at all. Yeah, I was about to say like Rodan mentioned, because I've also been part of your labs VR stuff, VR experiments at some point, you know, it's not real, you know, you don't believe it's real, but then kind of feel it. You feel the emotion, you feel that something bad is going to happen to you if you get off the plank. Even though you know it's not real, you know that you're in this safe little room. with nothing around you. So if you fall, nothing's going to happen. Yeah, I think this is totally adaptive. I mean, imagine if like, I don't know, way back in the past, there's some leader who tells everyone to like, trust me, I've given you the ability to fly, go run off that cliff. Like, it's adaptive to not believe that person, right? It's adaptive to not like let, to not have the kind of top down control. that could compel you to do that. Of course, we have examples like what happened in Jonestown and stuff that show that this can still be overcome. But I think there are some threats, things that are very evolutionarily ancient, like falling from a great height, that are really deeply embedded in this kind of like this fear response pattern. this aversive avoidant response pattern. So I think it's a good thing that this happens. You touched there on something I did wonder about reading through the study earlier. This idea of top-down processing versus bottom-up processing as it relates to, yeah, there's kind of the association between the physiological prediction of presence and then the subjective prediction coming from like the, oh, coming from the like emotion space. there's a general consensus that our emotions are interpretations of our environment and our physiological state on a very basic level. That's probably roughly what my understanding is. And so I wondered a bit about maybe there's some of that going on, because it seems like, I might have quite might have slightly interpreted your findings, but It seems like there's some sort of top-down where there's a lot of negative emotion. People are experiencing a lot of negative emotions. Where there's also that difference, what was it, with anger, which was also significantly increased across, or it would have been just been study one, I'm not sure if you can remember, but people feeling more angry and more sad and more fearful, just generally more negative. And I wondered if, but physiologically, this isn't necessarily relating to presence. I wonder if there is, I don't know, something going on there where people are understanding that they, this is a fearful environment, but I'm also kind of physiologically, I'm safe. I know that this isn't maybe, yeah, perhaps I know this isn't real. So my physiology will change absolutely, but not to the same extent that the subjective ratings. great, bring that back to prison. So I don't know. I wonder if there's, yeah, it's a top down processing stuff going on. Yeah. There's so your question touches on a, a good number of different perspectives on emotion. First of all, um, you talk a little bit about like dimensional views of emotion, like emotions being different on arousal and valence and motivational levels. And about the idea that there are these bottom up and top down. Influences and how we kind of construct an emotion on the basis of all of these different incoming signals and sources of information and about our expectation. And I want to say that, yeah, I, I do subscribe to that view of emotion. So I would describe myself as a constructionist emotion or I subscribe to constructionist theories of emotion. So like Lisa Feldman Barrett is the person who I think, uh, I probably attend the most to her writings about what emotions are. And actually I, I do think that our findings do speak to this relationship between bottom up and top down, um, in the sense that the theories of, of presence and how presence is related to emotion is that there are that emotions are associated with these changes in arousal. Um, so that's bottom up, right? We're literally feeling it from beneath our brain upward. And that would be directly informing presence that our ability to feel that change, which is just hap, which is caused by a separate thing happening over here, which is our emotions, you know, our emotions are changing our, you know, our part of our emotional changes, our arousal, our arousal is changing our presence. Well, that would imply that presence is very strongly driven by like bottom-up influences, right? But actually that's not what we find. What we found was that the bottom-up influences, they didn't have any predictive potency. So if we looked at the change in how people's heart rate like on the ground floor compared to at height, if we look at that change, so if we calculate the difference score, right? For measures of skin conductance and heart rate. Um, neither of those had any predictive potency in predicting how present somebody would be on the plank when they are afraid. So we just did some simple regression models to show that there was, there was just no productive potency of those models. I mean, like P of 0.9, like so not even close to being significant, but in both of our studies, we found that there was this relationship between subjective changes. So this is, again, another different score in fear ratings from below to when they're on the plank at height. And this was only in the height condition of study two. It was also found in study one. And when we threw it into a larger multiple aggressions model, we found that the d- subjective, the change in subjective fear was a significant predictor of presence on the plank. And then we did a mediation analysis that was able to show us that in that it was actually the change in the subjective fear that was explaining why people who were in each of our conditions had different ratings for presence. So like, if we look at, okay, are you in the height condition or the ground condition? If we just look at like that binary predictor, and we try to predict presence ratings, it significantly predicts it, of course. People who are in the height condition were more present on the plank than people in the ground or the control condition were. But when we enter in as a mediator, that change in fear ratings, it absorbed, it accounted for so much of the variance. that there was no longer a direct relationship between condition and presence. It was fully accounted for, fully mediated by that change in subjective fear. If we did the exact same thing with changes in physiological arousal as our mediator, so the exact same very simple triangle-shaped mediation model. There was a relationship, of course, between condition and presence like there always was, is not at all accounted for by changes in these physiological variables. So how does this relate to top down, bottom up? What I think this is saying is that these theories of where presence is coming from, they aren't getting at which aspect of the emotional experience is actually driving this sense of being present in your body, in reality, in the moment. And it is those top-down kind of influences that we really need to be attending to. When I say top-down, what I'm referring to is, how is it that somebody is coming to a subjective idea of what emotion they are feeling? Well, they're using their appraisal of the environment, that appraisal of their situation of the environment is driven by their experiences. You know, sometimes it's has, there are cultural effects that play into this there, but there's a whole different set of predictors, um, that contribute to what you are going to describe as, you know, eight out of 10 fear compared to the kinds of predictors of, uh, you know, interoceptive feeling of, Oh, my heart rate is increased. That's driven by a, uh, some, some different, some different causal. components. And so this is why I think it's important if we if we're trying to understand what presence is and where it comes from. I think we are, our theory so far have said look over here at these physiological signals. And I think that what our study is showing, what our studies have shown, is that we're looking in the wrong place and that we have to do the perhaps much more difficult but necessary work of trying to understand how different people come to the subjective emotional appraisal that they form. And it is through understanding those variables that we might be able to understand how presence is being driven by these influences. So yeah, I think that it, yeah, so hopefully, maybe that answers the question. But I think there are some other questions actually in your question that were. that I remember thinking, oh, that's actually a really good secondary point I want to come back to. But now, of course, I've forgotten. So hopefully we'll come back up. There was a lot which you said there in itself, which I feel can also could also get a little confusing. Like I lost I lost track a little bit because we were talking about so much within that. Could we just talk about the mediation for someone? who has less of an idea of what you mean by something mediating this relationship of presence. Yeah. Totally. So yeah, so just as a quick primer to what mediation is all about, in mediation, what you're trying to do is you are trying to explain a relationship between two variables. So. First of all, mediation is all about using relationships or correlations. There may be these diagrams and there may be these special statistical softwares that will run these mediation models, but conceptually, you're really just dealing with correlations between two variables at any time. So let's say you have a relationship between variable A and variable B. And variable B is the thing that your study is really interested in measuring. And variable A is like the way you've split up your groups. And so you go, oh, there's a relationship. Well, that that's good. That that means that my study, um, that my manipulation had an effect, that there is a difference between my groups on this measure variable B based on what I did in the study variable A. So that's, that's good. So in this case, it's basically on the ground versus the plank on the 80-story. 80-story is high, and the presence is what you're trying to measure. Yeah, so we needed that to be true. Yeah, yeah, yeah. I really don't even need to abstract it, probably. You know, I mean, exactly. So variable A is, do they go at height or are they on the ground level? And variable B is a... is how present they were. In order for any of these other questions to be asked, I have to show, first of all, that there's a difference. And so if there's a difference, there's a relationship between these variables. And so that relationship, you can create. Ultimately, you can do it with a multiple regressions approach if you wanted to. But I don't know if that's going to make it any clear to people who might not have a very strong statistical background. But suffice it to say that in the same way that you can do a correlation analysis with just kind of two variables, you can do a similar kind of analysis, but you can have more variables. And you can enter in a variable and say, OK, do these numbers in this what's called a mediator variable, variance in common with the numbers that are used to look at the relationship between variable A and variable B. This is called shared variance. And when this variance between these variables is shared, what we describe that as is, oh, this variable is explaining the relationship between these two variables. Now, this will always be the case. Like if you put random numbers, there will be some shared variance. But it's only in the case where there's enough shared variance that it would be considered a significant mediator. Oh, it's explaining so much of this variance that it is significant. And so that's what mediation is really doing here. It's some. or that's what we're really showing in this mediation model is that this variable that we put into the model, which is the change in how much fear people said that they were experiencing at time one and at time two. So like, let's say they said, oh, I'm feeling a two out of 10 fear when I'm in the start of this experiment. And then they go up on the plank, let's say, and now they're saying, nine out of 10. Okay, so that's a seven. That's a change in fear ratings of seven. That's their score. If that difference can account for the ways in which people's presence were different, depending on which condition they were in, that would be a significant mediator of the relationship. And so we found that But we found even further something called full mediation, which is sometimes that shared variance or that variance that's accounted for by the mediator variable is so large that when you look at the remaining numbers that haven't been accounted for yet, and you basically do the exact same correlation that you did in the first place between condition and presence, you can see if there's still a relationship between those numbers and there wasn't. If there isn't, then it means that the only reason that there was a relationship between variable A and variable B was because of the mediator variable. When that happens, what we say is that there's full mediation or that the effect is going through the mediator variable. That's how variable A is affecting variable B. And that's how you get to questions like causality as well. Like. things that are causes and effects as opposed to just relationships. Yeah, so just to bring it all home for your study, I guess one of the main things you found is that this explained variance is through subjective feeling and not through any of the physiological changes. Precisely. Yeah, exactly. Can I ask then, let's what What role you did kind of say like the physiological changes you measured with the heart rate were skin conductance and things of those nature. I would have thought that you would have seen differences between the two at ground level versus that level of versus yeah on the plank and because of fear and I thought that would have raised presence and you see none of it. So do you what explanations do you have for? I'm so glad that you're asking this question after that long winded explanation about mediation, because if you followed everything that I've said so far, then I think you can appreciate this because you're absolutely right. There were huge changes in heart rate and skin conductance. People were extremely sweaty on the plank relative to how they were when they were in VR on the city street. And that was, there was, there was a significant increase even in the control condition, but There are some figures in the paper that show how much more massive that effect is when you go to height When you're when you're walking the plank at height same with heart rate So there's a there is an increase in heart rate in the control condition from you know Being on the ground to being on the ground and on the plank, but in the height condition that change is pretty huge It's like 10 to 15 beats per minute of an of a change in your heart rate Which, you know, if anyone, if anyone's ever like worn like a Fitbit or one of those like heart rate monitors while they've gone jogging, like, you know, 10 to 15 beats per minute is like the difference between like, I'm not yet going for a jog and I'm getting started on my jog, you know, the first five or six minutes, you know, and then of course, you know, you get into a, a longer jog or a faster run and it'll go even higher than that, but we're not making them do much in the way of movement. We're making them walk three meters in one direction. Turn around, walk three meters in another direction, turn around, and then walk three meters, balancing on a plank. It's really physically not very challenging. So you're absolutely right. In this case, there was a relationship between variable A and variable, yeah, condition and presence, of course. The same relationship we found before. And it was not accounted for by the change in heart rate. or the change in skin conductance, despite there being an enormous change in both of those variables. Also, those variables were related to each other, which you typically would expect in a, you know, like heart rate and skin conductance. These are both kind of your physiological condition. They often tend to be cohere stronger with each other. They tended to be, they were, they were related to each other positively. And in one of the studies, however, we did not replicate this. the change in fear was related to the change in either of those physiological arousals. So, however, that was an effect that was only just barely significant and it didn't replicate, but it was trending in that direction in the second study. This is not very uncommon that emotional responses across different systems, so like subjective to physiological, will not be as strongly correlated as two measures within the same system. So two physiological measures or two subjective measures. But yeah, that's, it's kind of surprising that you would not see heart rate changes and skin connections changes having this, you know, predictive potency, if you will, that the subjective fear ratings have, because they did change in the direction we expected. The other one which I was thinking of, which I'm probably more surprised about this than I am about the heart rate and skin conductance, is you said you also looked at breathing. And I feel when people talk about, not well-being, what's the term, when they talk about meditation and breathing exercises and all that to be associated with being present in the moment and controlling your breath. relaxing and those kind of things. I would have thought that breathing in one some way would have been associated with presence when you feel you're breathing heavier and things like that. But that's more colloquial kind of discussion kind of thing where I thought that breathing might actually play a role as opposed to measurement. Yeah. Well, so it's a yeah, we didn't actually use... So it's one of the things that we can collect but we didn't use it as a dependent variable. And part of the reason why is because When we are collecting these physiological data, I mean, we're collecting it the whole time, but we're gonna select time windows. Like we're gonna only select what we call like epochs in the physiological data stream. And that's a very important design consideration. Okay, when are we going to do that? We decided to do that while they are. giving their verbal ratings of the emotions they're feeling. Why? Well, emotions are very dynamic. And if we're going to look at relationships between like these different subjective things that are being measured at the same time, because we're well, as close to the same time as possible, because they're going to like give a rating for presence, and then they're going to give a rating for fear or order reversed, then we should also take physiological samples from that same time. The problem with this as it relates to breathing rate is that we are getting a measure of breathing that is while they're speaking. And so their breathing rate is influenced by the fact that they are also providing verbal information. And so it's kind of prone to all sorts of extraneous variables, like the volume someone naturally speaks at, or how long someone might take to respond. And so. We didn't feel like we could really account for that as well. Yeah, that is one thing I did. Um, wonder about having seen some of the other research out of your lab is I've seen a lot of this kind of epoch related, are you saying a small window? Um, I've always been far more a fan of just probably taking the entire, like that entire, not just while they're taking the ratings, but like while they're on the plank for that entire period. saying we can get that entire section and then all of the time down the ground, which obviously it's a slightly different analysis and requires it. Yeah. So I do have tools to analyze it. But I just wonder, yeah, if you, I guess, I guess I wondered like, by the time people are giving these ratings, whether they maybe speak to maybe also. trying to calm themselves a little bit or appraise the situation so that they can give these ratings. Or if you just think, yeah, perhaps those physiological changes might have been dampened by the fact that they were giving the ratings. I'm not sure. Yeah. So in terms of regulation, which definitely is something that we took into consideration here, we had to kind of rely on the principles of random assignment in You know, one would just hope that by randomly assigning people to the height condition or the control condition, you're not systematically more likely to get, you know, people in the plank, um, the heights, uh, scenario who are all really good regulators. Right. And so you would, you would get like some people who are good regulators, some people who are not good regulators, like some people, and, and that is, I mean, I can't know how well somebody regulated in this study because that would imply that I know how they usually regulate and how this is how they did in this relative to that. But what I can tell you is that we observed, and I can say this myself because I ran the data collection in this particular study, the guy was the one watching the actual participants. And I can say that these participants were quite afraid, quite cautious. They, their movement was extremely slow and, um, and shaky and often, or rigid sometimes, um, and, but there were a, and then there was something about maybe like 10 or four or 5% some small percentage of people in the sample who had to stop before the, before they got to the end of the playing, like they were just too afraid, uh, and within that four to 10%. Um, A one or two people didn't even stand on the plank. They just, they just, as soon as I saw it, they were like, and that's the end of the experiment for me. Thanks. Um, and then there was another very small, but you know, non-zero percentage of people who didn't find it very frightening. You know, their fear might gone from a one to a three. Um, very, I don't think there was anybody whose fear ratings were like a one and remained a one. I actually don't think that. And I've gone through the data so many times. There's only 65 people or so in the study. So, you know, after a while you get kind of familiar with like every row, really. And I'm quite confident that was never observed. My point is though that you do get this nice distribution of responses where there is a nice. Well, there is this nice center, this nice bell curve that you're getting. And so, you know, it might be the case that there are some regulatory mechanisms going on. But in our case, as long as those aren't differing as a function of condition, you know, I see it as natural variability in behavior, you know, I see that as perhaps just part of the package of what we're trying to explain, you know. That was really, really well done. And we have so many more questions to ask. But I think because we should also think of wrapping up as well. Let's think, let's go into the more discussion section, speculative section of things. And let's, yeah, and let's talk about, you know, what you think are your reasons for what we've seen, or what you've what by we've seen what you've found, what are your speculative reasons for? We I know we've touched upon some of them already, but I guess this is your chance to kind of go and expand on your discussion, what you think is a good thing to go next. towards and yeah, I'll let you take the floor. So there are two things I, I want to say, um, in terms of the future directions here. First of all, I, I want to, I want to say that in psychology in general, the things that we're interested in trying to observe in the laboratory were not made to be measured. Like that wasn't one of the evolutionary goals that they aimed to serve. And oftentimes we encounter, uh, we, we encounter these, uh, problems with our theories where the, the things that we would need to do to, uh, look for evidence supporting or refuting those theories are really difficult to measure. But that doesn't mean that that's not. Necessarily true because. evolutionarily, like we weren't developing systems so that we could observe them in a lab. So, when I, what I think, and this has come through, I think, in a few of my responses, I think theories about presence need to be updated to incorporate the subjective appraisal element of an emotional response and how it is the subjective emotional experience that is driving how present we feel. And that might be true of other subjective, you know, things that may or may not be emotions. You know, of course, other emotions, I think that would, they would also have their own kind of relationship to presence. But we have to figure out how to get to this subjective appraisal. It's very tempting to want to look at changes in the body because we can look at changes in the body. But when it comes to like measurement of subjective, appraisal. We have to get this from explicit verbal or written or subjective responses, which are so open to experimental effects, leading questions, demand characteristics. It's a much messier thing and you really can't get into things that are very specific and And have the same kind of confidence as you would by looking at like heart rate variability or something like that, where, you know, people can't be trying to influence that deliberately, you know? I mean, they can, but, you know, not, not to the same extent. And so I think, first of all, that theories need to take on this difficult challenge of, of understanding how subjective emotional appraisals are contributing to presence. But in the meanwhile, I'm actually not done trying to test the theory as it is because I did this in one way, but this theory of presence that I've been alluding to is actually a predictive coding model. And if you've never heard of a predictive coding model, if anyone in the audience does not know what that is, basically what it means is in the case of presence, that we feel more present when the incoming information about the world, whether that's from our body or our senses from sight, smell, taste, hear, touch, whatever. When those things are predictable to us, when we were able to expect what was going to happen, and indeed that is what happened, that's when we would feel more present in our body. Like, oh, reality is operating the way that I thought it would. I must be in reality. In my study, one limitation of my study is that there is nothing that's really unpredictable. about, I mean, sure, this is not a very normal location to be in, walking along a plank, but your fear response should not come as a huge shock to you. You know, it should not be like, oh, I can't believe I'm feeling this way. This is not my body, or I am not in this world because if I was, I would feel a different way about it. So what I'm interested in doing right now is actually changing things like physiological arousal, not by using emotions. Not by using a context that has emotional value, but by directly changing physiological arousal in ways that either can or cannot be predicted. And if this theory holds, then when the changes in physiological arousal that I can cause directly, for example, through like, if I get people to drink coffee or something like that, um, but like in VR, right? So there's in like a non-emotional VR and they're drinking coffee and I asked them how present they are. I would expect that if the physiological change that's happening is less predictable, that they would be less present or they would report feeling less present. But that is kind of what I'm trying to approach in my next, my follow-up line of studies. So that's two ways forward that I'm charting right now. I think I was wondering earlier, can I discuss it or not? wondering about, I touched on in my long-winded question about kind of the negative emotions and their role in this more so than just fear. So wondering about like, did you run these mediation analyses? Well, one question, do you run the mediation analyses with other negative emotions? Secondly, it's because this seems so much like this general negative emotional response I wondered if perhaps a different measure of those negative emotions. For example, PANAS might have been the Positive and Negative Affect Schedule, might have been also something worth considering implementing, although it is a rather long measure. Yeah. Um, so this is actually a really, it's interesting that you're asking me this question today, because actually one of the things that I'm going to do right when I'm done with this is, um, I'm, I'm actually going to be recoding a lot of the emotions based on, uh, positive, negative arousal, I mean, uh, surveillance and then high and low arousal and, uh, and, and these different dimensions and I am going to run some of these regressions, um, but not. For this study as such, I'm actually, it's for a new study that has to do with whether or not large language models know how people respond in emotional situations, which that's a whole different, that's a whole different can of worms, which I would be delighted to talk to you about, but probably not until I have an answer, which I'm very close to having an answer on. But with your question about how much is this just negative emotions or, you know, as opposed to fear specifically. I actually, I don't feel that it is the case that we induced just a negative emotion or just a high arousal emotion. And the reason why, and I'm looking at, in my study, it's table two, there are a number of different t-tests comparing the average emotion rating at time point one and time point two. And you'll see that like they all significantly changed, anger change, sadness change, happiness change, relaxing, and they all change in the direction and in a common direction. So anger does go up, sadness goes up, fear goes up, but it's the magnitude, it's the effect size here that I think tells you the story better than the significance. So sure, P values very, very low in all cases, but for fear, the Cohen's d, is in table two is 2.18, which is pretty gigantic. I mean, a large effect would be a Cohen's D of like 0.8, right? So this is like more than twice the size of what would be conventionally considered a large effect. But then if we look at the change in anger and sadness, the Cohen's D is 0.46 or 0.43, which is more like a moderately sized effect. So I think the scale of the change really puts it in a different category that makes me feel Sure, I'm not saying that we didn't induce a negative effect overall and that people weren't a little more, you know, sad and angry as a result of the, like, feeling this, this fear. But I think that, you know, if you look at, yeah, fear and anxiety both have a co-initiative greater than two. I think that shows that we're seeing things that are much more in a kind of specific emotional direction or pointing to a very specific point on these dimensions of emotions. than just being left of the... If you imagine this as a negative positive, it's not just all these things on the left are being increased in the same way. Yeah. Cool. And for those who are wondering about the numbers and tables, I will link the paper in the show notes and description and everything. And if possibly even put up some of put up images of the tables and figures, which Chris mentioned along the way as well in the video. Hopefully I can do that. But yeah, so that you guys can also take a look at everything. I encourage people to read it, read the paper for themselves. If I can put a very final note in here, part of the reason why this paper took me so long to write, it took me a very long time to write, is I really wanted to write a paper that was for somebody who was getting into experimental design or trying to understand a large experiment, but didn't need to have like a huge background in or an expansive expertise. in stats and modeling these different various effects. I was trying to write this for people like that. And so I'm sure there are a couple of places where I say something where I wish that I could have expanded a bit more. But if you're somebody who's interested in psychology, but you're in your first couple of years of uni, or you have a background in it, but not this area. I've tried to write this paper in a way where you do not need to have a lot of background in this in order to be brought along. So it's a long paper. There's a lot of analyses, but I try to hold the reader's hand through all of it is what I really want to convey. No, awesome. Thanks so much for that, Chris. Thanks for coming back on. Thanks for writing. Cheers. No, that was really good. Like last time again, if people want to reach out to you, they can just email you, contact you. They may. Yes. Email is always open. And you had given one piece of advice back then when you had come, but I do ask people to give our audience members one piece of advice to leave with. So if you do have anything else to share from two months earlier? uh advice to um well i guess it depends on to whom uh do you mean like about this something that would be advisable as a result of kind of this paper or to somebody in psychology who's i guess i guess general advice to our listeners Okay. Um, from you general advice to our listeners is to, uh, I don't know. I'd say, you know, I would like the, I would like the world to start thinking more about this idea of how present they are, uh, in any given moment. Um, there's a lot of times when I ask people about presence, it seems like they haven't really considered this before this aspect of this dimension of their existence. And. I'm not saying that I think there are health benefits doing or maybe they are, but I, I would love if people just started to think more about how present they're feeling in any given moment, because I think there's so many relationships between the environment and presence out there that we've never really identified because it's not part of the discourse yet. So. Attend to your presence is I guess my advice and, um, and if you do discover anything that is a. that you think is interesting. Yeah, my email is open and I look forward to hearing from anyone. Awesome. And yeah, thanks again, Royden. Thank you so much, Chris. Thank you for having me. This has been a delight. Thanks everyone for listening and until next time, take care.