WHAT VR CAN SHOW ABOUT HOW HUMANS NAVIGATE NEW SPACES

Cassandra Engstrom is beginning her second year in the M.S. Program in Cognitive Neuroscience at the Graduate Center. After more than a year of working remotely because of the COVID-19 pandemic, she has resumed in-person research and experiments in the lab of Presidential Professor Tony Ro (Psychology, Biology). While at the lab in early September, she filled us in on her project and how it feels to be back at the Graduate Center. 

The Graduate Center: Tell us about your research. 

Engstrom: I am currently conducting a study in virtual reality to test what information humans most willingly rely on when navigating through new spaces. Most of the navigational strategies used by mammals involve building a mental model of the environment that remains relatively constant. When they know the structure of the world, they can plan what they will do before they do it. But this takes time and energy, and it takes up space in the brain. There are other forms of navigation that are more efficient from a memory standpoint, but they only really help an organism get from some point A to some point B along an inflexible trajectory. Humans probably use both strategies — which one they use will depend on the situation they’re in. 

I’m interested in what happens to the preferred strategy when the clues we gather from our experiences, for example surrounding visual patterns, inaccurately suggest where we are positioned relative to the mental maps we are using to navigate. In other words, I'd like to know how we deal with conflicting information. Participants in my experiment, who are essentially moving through a virtual city, usually tell me they get rather “disoriented.” As long as they are not suffering from this, I’m happy to hear that! It will make for a very rich dataset, I hope. 

GC: How does it feel to be back at the Graduate Center conducting research?

Engstrom: I feel very grateful. After a year of mostly solitary project development and a lot of debugging, it feels good to finally test what I’ve spent so long creating. It’s also very interesting to get feedback from my subjects who each describe a very different approach to tackling the tasks l give them. 

GC: Are you taking in-person classes as well? If so, what is that like? 

Engstrom: Yes. I’m actually taking a graduate course in machine learning at Hunter. It’s a small class, which I think is excellent for group work and more hands-on skill building. The other members of my class are in the computer science master’s program at Hunter so I’m hoping to take advantage of their preexisting knowledge base in case I get desperately lost, which is not far-fetched. The material will be a challenge for me, but I think it will be easier to absorb in person. 

GC: You have a B.A. in Architecture with a minor in literature from the Cooper Union. How did you decide to get an M.S. in Cognitive Neuroscience at the Graduate Center?

Engstrom: To me, one of the most interesting challenges of architecture is distillation. You have to distill very complex structures that often exist first in your imagination into a notational language that is precise and interpretable by others. That’s kind of what the nervous system does, but in reverse: All of our subjective experiences arise from the collective activity of billions of interconnected cells that also use a precise and unambiguous language with one another (much like architectural notation). Of course, no one “speaks” the language of the nervous system, so we have to infer things about that language through rigorous experimentation. There is also this idea of “latent” features. The drawing sets that were used to construct a building are not actually accessible when you’re inhabiting the completed work — that information is latent and probably only derivable by an expert. There are also latent features that underly our visual experience. Whenever we perceive something, for example a wonderful painting, our brains are responding to features that we as conscious beings do not, and probably cannot, discriminate. I guess I’ve always been interested in that kind of translation between different representations. In the case of architecture, it's between instructions in different dimensions and their collective product. In the case of cognitive neuroscience, it’s between brain physiology and its phenomenological outcomes.

As for my minor in literature, it sort of happened by accident. I like to read and write, and it’s important to shake things up so you don’t operate in one mode all the time.

GC: How’s the program going for you? 

Engstrom: I have really enjoyed it and I think there’s something for everyone. The independent research component has been particularly valuable for me, but the faculty and core courses are awesome. I’ve learned a lot. 

GC: What are you hoping to do next once you have your degree? 

Engstrom: I plan to pursue a Ph.D. in Computational Neuroscience.

Published by the Office of Communications and Marketing.