Subscribe
Share
Share
Embed
In this episode of The Highlights, we're joined by Yeon Soon Shin, who completed her doctoral degree in Neuroscience at Princeton in 2020 and Rolando Masís-Obando, a 5th-year graduate student in Neuroscience. We discuss their paths to graduate work in psychology, the two distinct VR environments that they created, and the importance of using VR in their research on the environmental reinstatement effect.
Brains, Black Holes, and Beyond (B Cubed) is a collaborative project between The Daily Princetonian and Princeton Insights. The show releases 3 episodes monthly: one longer episode as part of the Insights partnership, and two shorter episodes independently created by the 'Prince.' This show is produced by Senna Aldoubosh '25 under the 147th Board of the 'Prince.' Insights producers are Crystal Lee, Addie Minerva, and Thiago Tarraf Varella. This show is a reimagined version of the show formerly produced as Princeton Insights: The Highlights under the 145th Board of the 'Prince.'
Please direct pitches and questions to podcast@dailyprincetonian.com, and any corrections to corrections@dailyprincetonian.com.
Thiago 0:00
Hi everyone. My name is Thiago and I'm a grad student at Princeton University and I am your host. The Highlights is a sister podcast to Princeton Insights in collaboration with the Daily Princetonian. Insights is a newsletter written by Princeton undergrad, grad students, and postdocs. We write about the most exciting and groundbreaking research being conducted here at Princeton in the form of short, fun, easy to read reviews. We cover a range of topics including psychology, neuroscience, biology, computer science and physics to name a few. Make sure to check out our website at insights.princeton.edu. Right now, I receive my fellow graduate student Paula Brooks. Say Hi, Paula.
Paula 0:38
Hi, everyone. I am super excited to co-host this episode of the Highlights with Thiago. And I recently wrote a review for Princeton Insights about the paper we'll be discussing today. It's called "Context dependent memory effects in two immersive virtual reality environments on Mars and underwater", and it was written by Yeon Soon Shin and Rolando Masís-Obando, who are both joining us today, along with other people from the Norman lab. But before we get started, we're going to introduce our amazing guests. We'll start with Yeon Soon. So Yeon Soon is a postdoc in the Crockett lab in the Department of Psychology at Yale University. She received a PhD in neuroscience from Princeton working with Yael Niv and Ken Norman, and she uses computational models to understand how human experiences are organized in memory. Hi Yeon soon. Thank you so much for joining us.
Yeon Soon 1:39
Hi, thank you for having me.
Thiago 1:41
Thank you, Paula and Yeon Soon. Rolando Masís-Obando is a graduate student at the Neuroscience Institute at Princeton University, and is currently co advised by Ken Norman and Chris bazzano. Broadly, Roland is interested in how contexts of whatever shape or form influence how we perceive and remember events in our life. He enjoys the intermingling of his interests in filmmaking and art, with all of his science work, and often finds they feed off each other synergistically. Thanks for coming, Rolando.
Rolando 2:13
Thanks for having me.
Paula 2:15
Yeah, so so let's just jump into it. To start off, can you both tell us a little bit about how you ended up where you are right now?
Rolando 2:24
You know, originally, I kind of wanted to do physics, actually, when I was in middle school, for some reason, I thought that, you know, like, physics was like, this amazing field where we could ask all the really tough questions about you know, existence and humanity and how things work. So I was really always wanting to ask, like, you know, why, why is this work this way? Why, why so, you know, I do like science fair experiments and all that. But then the high school, you know, under the influences of the matrix, and all these other movies and all these other things, you know, I was like, Oh, the brain, the brain. So I started like, looking into like, why the music, you know, affects the brain like searching, searching things up. And was like okay, yeah, I think I want to do neuroscience. So then, like, from from there on, I was okay. I feel that to do neuroscience, it still is still like this physics kind of mentality of I think I need to know the nitty gritty. So I wanted to do neurobiology, learn about bacteria, learn about all these other, you know, biological organisms and make my way into the brain. And then from the brain, you know, I don't know solve the world. But obviously, that's not happening. But what did happen is I went through biology. And then I moved when I went to grad school moved into cognitive neuroscience, where I could ask the more higher, no more conceptual, more abstract questions. And still, I could still ask the why and how, but with humans instead of animals. And yeah, so I'm kind of just living the dream, I guess.
Paula 3:42
That's cool. Awesome. How about you Yeon Soon?
Yeon Soon 3:46
Yeah, my path was slightly different from Rolando, I would say because in my case, I was always interested, interested in psychology. And just like, I wanted to understand myself and people around me like how people work. And I majored in psychology in college. And it was it started as more like counseling psychology. And as I was studying, I was like, Oh, actually, what I want to understand is like the nitty gritty details of like how the mind works, it's not just applying what's already known. So I moved into cognitive neuroscience, and I did masters in Korea. There I started, like learning about reinforcement learning models, and really got fascinated by how much it can explain how the mind works, and also how that gets connected to behaviors. So for grad school, I really wanted to do computational modeling in more depth, and that's how I ended up in Princeton and did computational modeling. Ultimately, my goal is to understand like a lot of social things that like, for example, like how people stereotype other people. So for a postdoc, I am working in a social psychology lab, that's, that's one step closer to my ultimate goal
Paula 5:14
Cool.
Rolando 5:15
And your ultimate goal is to rule the world.
Yeon Soon 5:21
That's right, Rolando. That is my goal.
Thiago 5:25
So could you explain a little bit how was this collaboration born? Like, how did each of you decide what to do and who had the idea of working together?
Yeon Soon 5:34
So I've been working on this general theme project for a long time when we started this collaboration, and I used to use video games to manipulate contexts. And as I was running many different experiments on this study, like the previous iteration of this involve, like zombies throwing fireballs, and things like that. And it wasn't really working. And around that time, Rolando was doing virtual reality experiments in the lab, and the video game had potential to become virtual reality setup. So yeah, we started talking around like 2017 18, something like that. And yeah, we started building these two very different worlds to investigate context dependent memory.
Paula 6:29
That's cool. So I guess transitioning from the collaboration and getting a little bit more into the paper, can you tell us why we should care about your paper? I feel like I'm giving you all the hard questions.
Yeon Soon 6:45
So why we should care about this paper, in our life, you know, like context plays a big role in how we organize memories and how we retrieve information in the right moment. The contribution of this paper I would say is that we really try to maximize the conditions, to maximize the features, that leads to that effect, because it's been very tricky to show that effect in the lab with virtual reality. And with all the manipulations that we did, we could get the robust effect. And we suggest a lot of features that we think is important, especially representing the mental context, as opposed to just spatial context.
Rolando 7:31
Yeah, I think the young sees exactly on the on the target. Context is like this super important thing, right? It's, it's everywhere you have like, I guess, like the maybe the best example to use here is, in the one where you actually use in the intro of the paper was, you know, when you're going back to a childhood home, or you're going to somewhere in the past, or like some other context, you know, sometimes that context was powerful enough in some way to you that it all of a sudden, you have a bunch of memories flood in, right you if you're you go into some room, and all of a sudden you remember playdates with friends, it almost feels obvious that, you know, why do we need to use science to show that this effect exists? And like, you know, why is it so hard to show experimentally? But it turns out that it's kind of hard to show experimentally. And so as you said, like, yeah, we, I think we've provided like a good repertoire of features that would make this more, you know, accessible in the experimental space for people that want to keep for including us probing more and more and more of what it means to like, exist in the context and how it like actually helps organize our memories.
Thiago 8:36
You both mentioned that it's hard to test this effect, but why, it why is it hard?
Rolando 8:42
Well, well, there's many many reasons when the so the main paper that we use is like the the kind of main inspiration for this is, is a 1975 paper by Godden and Baddeley and this paper, you know, which was like super, you know, seminar in the sense that tons of other papers started like building theories off of it, you know, starting to make trying to replicate and in some way or form, in this paper, they had divers, people that, that do scuba diving and all that, learn words underwater, while they were, you know, you know, breathing underwater, and then learn words on land. And like, again, the context memory effect, right is this effect where what you study in one environment, if you get tested for that same knowledge in that same environment, then you have good memory for it. But if you get tested in a different environment, not the one where you did the original learning, then your memory not might not be as good. To us when they give you the example of the Salika or like, you know, the memories coming in when we go to childhood home, it seems like Oh, but that seems kind of obvious. But the experimental paradigms that people have used to kind of like, get at this effect, miss a few of the attributes that we actually ended up hitting in our own paper.
Yeon Soon 9:47
So for example, some of the studies that tried to replicate that effect, like use different rooms as different contexts, and they decorate these rooms in very different way like One room is like Red Room and the other room is a blue room or something like that. Switching from bedroom to Blue Room should mean something to the participants for the effects to come about. But in a lot of cases, that's not really necessary for the participants to even think about the different rooms as different rooms, because they're in an experimental situation. And this is just like psychology experiment building. And I do not need to think about Blue Room and Red room in different way, in any way.
Thiago 10:29
Like in their brain, both of the rooms are just the lab.
Yeon Soon 10:32
Exactly, exactly. So they are merging the context into one. And in that case, even though the experimenters are manipulating these two rooms as different conditions, they're actually not being represented as different separate rooms or separate contexts.
Paula 10:51
So in the title of your paper, you reveal your hands, so to speak, and talk about having these two contexts Mars and underwater and again, talking about the importance of context, how how did you actually use virtual reality VR to test this context dependent memory?
Rolando 11:09
as obvious as it might end up sounding we created literally two different worlds, like in like video game worlds, just imagine, like you're playing like some video game. And you know, there's one map and there's a different map, we had two different maps, one map was underwater map. And the other map was Mars map. And the way to make it, you know, as immersive as possible was to, you know, have different sounds for each environment.
Thiago 11:31
So Can each of you describe a little more about the environment, let's say, Rolando describe the Mars environment, and Yeon Soon describes the underwater environment?
Rolando 11:43
Yeah, so the Mars environment, you know, looks like a little bit like what you'd see in like sci fi movies, and what you'd see in you know, some Mars rover picture except very red. And, you know, you have like the grounds you have like these little base camp looking sci fi little centers. If the participants wanted, they could look up and there are a bunch of ships that were kind of just hovering in the air, and you'd see them kind of like moving up and down, if you squint your eyes, because they're pretty far out to make it more realistic, you could see across the horizon, and then you'd see like misty clouds, and maybe some shooting stars in the distance, you'd be able to teleport with a special device and in order to scan the rocks that contain the words, because at the end of the day, they were, you know, scanning war rocks to you know, with this particular gravity gun that we we had, you know, they click a little button, we have a little sound, and then they be you know, readings, aka scanning the rock to reveal the word. And so, yeah, so people would just like, you know, walk around, it was a pretty open space, pretty, pretty much free world.
Yeon Soon 12:51
So for underwater, conceptually, it is you're finding ruins from shipwreck or something. So like, there are many treasure chests, and you need to open the treasure chest to find the items. In the middle, there is a big treasure chest where you go for instructions. And so to open that big treasure chest, you need to find the golden key that's on the ground. So like you move, and you see that there are like some fish and some dead fish as well. That was our joke. And there were like really beautiful jellyfish just floating around and you see the bubble sounds and you open the chest with the golden key and you continue to find the more wars by opening up more chests
Paula 13:42
what you're describing, it's it's really obvious that it's very different from a red background on a computer screen versus a blue background, changing context in the lab in that regard. But I guess Can you elaborate a little bit about how is the VR context different from the real life context use in other experiments? And then what's the advantage of using VR?
Yeon Soon 14:06
Right. So the difference, I think, is that it is associated with the tasks that they need to perform. So the motions were related to the task, right, like they needed to find words and judge them. And that was dependent on the context, like the task itself was dependent on the context,
Rolando 14:27
in our context, and Mars are underwater, they had a sequence of things that they had to perform in order to, you know, get some gets a completion. And so by having those those tasks be different across environments, we're able to kind of really make each environment unique with its own world, both visually, orally, but also mechanistically. And that kind of helps separate both both mental spaces so that when we try to try to test this context dependent memory effect, we actually get to piggyback off of the strong associations that each environment has with itself.
Yeon Soon 15:03
So people going into the experiment has already some conceptual knowledge about what Mars should look like, although we've never been there, like, we know something about Mars, and we know something about what it should be like and underwater so that there is already rich conceptual knowledge about those environments. And that's, that's what we really want it to rely on.
Rolando 15:28
And you can't really do that, right? If you have like, some version of this on like, on like a computer screen. I mean, yeah, there are movements that you could do, but it's not as realistic as actual physical movements that are more immersive. And, you know, one of the things that we wanted to really do is, you know, we wanted to kind of like, try to include as many features as possible to make this as closest to reality, but controllable than it would be if we were trying to use two different rooms that look different, and like try to recreate Mars in one realistic room versus another. Like, in VR, we have so much control, because we can program everything. It's a video game, we can do whatever we want.
Paula 16:16
I'm wondering if maybe we could transition a little bit more to talk about the task itself. Like, you talked about people learning words, but But then what was the test? And then from there, maybe talk more about, were you able to get this environmental reinstatement effect using VR? Oh, I mean, yes, you are. But but just elaborate more more about that.
Yeon Soon 16:43
Right. So the task as Rolando alluded to, the task was to judge whether the item was helpful in surviving in this environment, because humans needed to find some newer environments to like to have it because the Earth was not inhabitable anymore. That was the story, cover story. So like, if you find, for example, a Jeep to be useful and surviving in Mars, then you should say that's useful. And if you think that blanket is not that important than surviving in underwater, then you should say that that's not that helpful. As they were judging these items, they didn't know that it was a memory experiment. So it was a surprise memory test. And the reason we did this is because once people know that it's a memory experiment, people are very good at making very elaborate strategies to remember these items. And we didn't want them to rely on any other strategies than then just finding these items to the context that we provided. The task was promoting them to bind these items to their context even more by, you know, thinking about the utility of these items, what we thought was what we hypothesize is that, when they say that, Oh, it's not useful, then those items should not be bound the context as much. And so these not useful items should not show strong context dependent effects. And that was what we found as well. So useful items show context dependent memory effect, but the not useful items did not show
Thiago 18:31
What sort of questions can this new technique with VR answer that we weren't able to answer before?
Jarome 18:37
All of them, I'm just kidding.
Thiago 18:39
And take over the world.
Rolando 18:41
You know, like VR, you know, it's just a tool, right? You know, you have, you know, you can use different tools to reach the same end. Some tools will allow you to get to the end faster or more efficiently, while other tools will answer different questions along the way. VR made it very useful for us in the sense that, you know, this is like a very, a hard to reproduce effect context may be more effective when it's like immersive and like, quote, unquote, overwhelming, rather than something on the computer screen. And so for us, this tool was quite useful for the ends that we wanted to reach, which was, hey, this effect is definitely real, we just need to find like, some set of properties that would be useful to kind of use to see if we can elicit it. VR is definitely one way to do it. It's not the only way though. For other types of experiments, there's a whole world of of science that could use VR, currently even use right now for like exposure therapy, you know, helping people with phobias. And so the, the space is very large, you know, might be a little bit unethical to some extent, right. So someone has a spider phobia. Let's bring an actual spider into the room. You know, that kind of exposure therapy is a little hard plus you have to control The actual spider and like, you know, you have to keep it alive or no, there's tons of other stuff, right. But if you are makes life much easier, it's not a real spider, but you can still elicit some of the effects that would occur with someone that has a phobia for spiders. So again, it's just a tool that depending on how you use it, you can do a lot of cool things.
Yeon Soon 20:17
And I'd like to add also that the measurements that we can get out of VR can be really rich, we only analyze the recall data. But we can also look at the locations and like how they were moving, what they were looking at in the moment and things like that. So it actually provides really, really rich data, which I'm really excited about
Thiago 20:42
Nice. At least for me as a final question from what I gathered, Yeon Soon, is your next step is going to a social psychology lab, Rolando is probably taking over the world or something. What would what would be the next steps for this project? Specifically? Do you have a follow up? Or is in your view, working on something that follows up are are your next steps kind of diverging from that?
Rolando 21:08
Yeon Soon and I have played a lot with the idea of follow ups. And we have like a lot of different pet ideas. You know, you know, as like Yeon Soon just mentioned, you know, VR game provided tons of rich data that we could look more in depth and find more connections with memory, and a lot of the properties of the space itself.
Thiago 21:25
Okay, yeah, that's pretty cool. That was an incredible conversation. This research is super cool. It's super interesting, both the result of the memory, the environmental statement effect, and the tool that you're using to test it. So thank you very much for coming here. And hopefully, I'll see some of this results published as well, those next steps.
Paula 21:52
Yeah, thank you for talking.
Yeon Soon 21:55
Thank you for having us.
Rolando 21:57
Thank you so much. This was tons of fun.
Thiago 22:00
This episode of the highlights was written by Thiago Taraff Varella and Paula Brooks. It was produced by Isabel Rodrigues under the 145th Managing Board of the Daily Princetonian. For more podcasts and other digital media from the Prince, visit www.dailyprincetonian.com. Many thanks to Yeon Soon Shin and Rolando Masís-Obando for speaking with us. To read more about their work, check out the Princeton Insights article covering their research, which can be found in the description of this episode. Thank you for listening, and until next time.
Transcribed by https://otter.ai