Teaching Statement
My name is Ilya Musabirov, and I am a second-year PhD student in the Intelligent Adaptive Interventions group, supervised by Joseph Jay Williams. In the group, I apply adaptive experimentation to design self-improving systems, helping researchers and instructors study and apply best practices in learning design and motivation.
This reflects my own teaching philosophy — focus on evidence-informed approaches to teaching, supporting student self-regulation skills and motivation.
Before starting my PhD at UofT, I was an industrial data analyst and then spent seven years designing and teaching courses and programs, introducing hundreds of non-STEM students of different majors to the basics of statistics, data-centered computing, and data-driven decision-making at HSE University St. Petersburg. During this time, I got three Best Instructor Faculty awards, multiple course awards “Usefulness for future career” and “New knowledge”, and the “Dream Mentor” Jubilee Teaching Award.
The most relevant part of my teaching and course design experience is co-design and teaching Data Science undergraduate minor for students of different non-STEM majors. Having a hard task of making it accessible (and useful!) to students with very different levels of computational skills and math backgrounds (from Design to Economics), I focused the design on providing quick and early feedback for skill formation, using small authentic tasks to provide an immediate sense of accomplishment and increasing course utility motivation, and simulation-based explanation to build intuition about stats and machine learning methods. These efforts led to the rise of career or grad school choices, bridging SocSci and Humanities with data skills.
One example of my approach to learning design is an original module introducing students to the idea of machine learning models via participatory simulation. During the simulation, students are tasked first with developing systems of if-then rules for a simple simulated data classification task (e.g. credit decision). Then they have an opportunity to submit and compare their rules with a discussion of a variety of quality metrics, starting with their own ideas on what good models are. In the next stage, they recreate the idea of recursive partitioning and are introduced to classification and regression trees.
While helping with emergent transitioning to remote teaching in 2020, I got to understand how large was the part of teaching success which relied on face-to-face communication and peer learning and how obvious became the need for intentional design of learning which would account for supporting student motivation, helping them with self-directing their learning, and dealing with issues. To focus my research on that, I joined the Intelligent Adaptive Interventions group at UofT.
Here I’ve been a TA for Intro to Programming and Human-Computer Interaction courses and led learning design in our group to help undergraduate research assistants to learn research stats and computational skills. The most recent example of my work at UofT is the design and improvement of a learning module for CSC428 (Human-Computer Interaction) students on Hypothesis Testing which helps students to build intuition and connect the main concepts – sample and effect sizes, power, observed differences, expected distribution under null hypotheses. This is done using a guided walk-through series of simulations and uncertainty visualizations, integrated with reflective questions and formative assessment.
In addition to university teaching and TA work at UofT, I developed five workshops and tutorials for professional audiences of different skill levels and domains, e.g. workshop in Data Visualization at Learning Analytics for Nordic Learning Analytics Summer Institute (Tallinn 2019), basics of AB testing for game analysts, Machine Learning for Social Science instructors (St.Petersburg).
I hope that my teaching and course design experience, learning science knowledge, and love of R will help me contribute to the success of DSI teaching initiatives.