Instructor Interview
Below, Prof. Josh McDermott describes various aspects of how he teaches 9.35 Perception.
OCW: Your lectures draw heavily on illusions (both optical and auditory) to explore the science of perception. Why do humans find these illusions so fascinating? And what do they reveal about the way our senses work?
Josh McDermott: Part of the fascination comes from the fact that perception tends to be hidden from human observers. We perceive an estimate of the world that is the result of incredibly refined and sophisticated machinery in our brains. But we don’t usually have conscious insight into the workings of this machinery, and because perception is usually effortless, the layperson might be unaware that there’s machinery there at all. Illusions make us realize that there are perceptual processes that are always happening under the hood, and that these processes are doing a lot.
OCW: What have you learned about lecturing in your years of teaching this course?
Josh McDermott: Summaries are very useful. I should probably give more of them (e.g., at the end of lectures, or at points of transition in mid-lecture). It helps people consolidate and confirm their understanding.
OCW: The Illusion Lab assignments, in addition to requiring students to research, plan, and implement an original illusion inspired by existing research, also require that students provide substantial feedback on their classmates’ illusions as well as feedback to the course staff on the assignment. Can you share your thinking on the pedagogical value of this kind of feedback?
Josh McDermott: First, I want to give my thanks and appreciation to a treasured former grad student, Maddie Cusimano, who was responsible for the introduction of the Illusion Labs into the class during the years that she was the class TA. It has been a nice addition. Regarding student feedback on the illusions: Part of the motivation is to provide the illusion creator with some informal data speaking to the extent to which the illusion “works” on other people. It’s a bit like running an informal version of an experiment that one might conduct to document a new illusion.
OCW: Your policy on AI, as of spring 2024, simply states that students “may not use ChatGPT or a similar tool to complete problem sets or labs.” Has it been difficult to enforce this policy in practice? And do you see AI tools as having any legitimate uses in a course such as this one?
Josh McDermott: Back then it was pretty obvious when students were using AI, and we just asked them to re-do assignments when we noticed it. We’re now in a different place. I think the next time I teach the course I will make problem sets optional—essentially, they will become study aids for (in-person, hand-written) exams. AI can be useful in learning new domains, but most of the evidence I’ve seen so far indicates that learning outcomes are hurt when students rely on it. It can give the illusion of understanding and prevent the struggle that tends to yield intellectual growth. I think teaching will have to evolve in the coming era to help ensure that humans continue to learn how to think despite having very powerful crutches that can accelerate our work in many ways.
Josh McDermott: “I hear “Laurel.”
Assessment
Grade Breakdown
The students’ grades were based on the following activities:
- 35% Problem sets and labs
- 30% 4 midterms
- 10% 4 quizzes
- 5% Attendance at Illusion Lab classes
- 20% Final exam
Curriculum Information
Prerequisites
9.01 Introduction to Neuroscience or permission of the instructor.
Requirements Satisfied
9.35 can be applied toward a Bachelor of Science in Brain and Cognitive Science or a Bachelor of Science in Computation and Cognition, but is not required.
Offered
Every spring semester.
Student Information
Enrollment
33 students.
Breakdown by Year
The class contained a roughly even mix of second- to fourth-year undergraduates.
Breakdown by Major
Most of the students were majoring in Brain and Cognitive Sciences or in Computation and Cognition.
Typical Student Background
Most of the students had some prior exposure to neuroscience and cognitive science, often from having taken 9.00 Introduction to Psychology and 9.01 Introduction to Neuroscience.
How Student Time Was Spent
During an average week, students were expected to spend 12 hours on the course, roughly divided as follows:
Lectures
Met 2 times per week for 1.5 hour per session; 26 sessions total; mandatory attendance.
Recitations
Met 1 time per week for 1 hour per session; 12 sessions total; mandatory attendance.
Out of Class
Outside of class, students completed assigned readings, worked through problem sets and labs, developed original optical and auditory illusions, and studied for quizzes and exams.
