Teaching Philosophy

Core Belief

If I treat teaching and learning as something to study, they do not behave like a simple natural law, where doing A always produces B. They behave more like statistical thermodynamics. Nobody can say where a single molecule will go next, but once the temperature and pressure are set, a large collection of molecules behaves in ways that can be predicted with confidence. Teaching looks like that to me. I cannot force learning to happen in any one student, but I can set conditions that make it more likely.

In last year's statement I described learning through the second law of thermodynamics, which is itself a statistical law: left alone, things drift toward disorder. Deliberate learning runs against that drift, and it happens only when students and I both put effort in. That belief held steady over the past year. This year I focused on three variables: who my students are, how I teach, and, newest of all, AI.

My Students

My education master's program taught me that community college students come from very different backgrounds. Teaching them showed me how wide that range really is. I have taught a student who took chemistry with us at high-school age and is now in a combined bachelor's and master's program in engineering at a four-year university. I have taught adults who already had careers and enrolled because they wanted to keep growing. Most of a section sits somewhere between those two. State policy widens the range further: since AB 705 and AB 1705, one section of CHEM V101 enrolls students whose math background runs from full fluency to foundational gaps.

So I plan a course for that whole range from the first day, using the methods in the next section.

Teaching Methods

My goal is to make learning reachable, not effortless. That usually means putting an intermediate step where the gap is too wide, such as a simpler example before a complex one, or a worked line of reasoning students can see before they try it themselves.

I also show the same idea in more than one form. For CHEM V101 lab there are 67 short videos of the experiments, and for lecture a library of 35 browser simulations I built this fall. At the bench, students get a checklist. For Experiment 2 this September, students ticked fifteen boxes while they worked, and every item on the next week's lab quiz traced back to at least one of them. The lab quizzes ask about the work done at the bench, because the lecture quizzes and exams already cover lecture content. The checklist also answers a point Malia Rose-Seisa made after observing my lab in 2023: students need to know exactly what to write down, and where.

I also keep records. Through the Math Readiness Initiative I have been recording the exact point where students stop in a calculation: at a fraction, a ratio, or a unit conversion. That turns "weak at math" into something specific enough to act on, including whether two students with the same score are stuck in the same place.

AI

Who my students are and how to teach them have both been studied for many years. AI has not. Everyone is a beginner, new models arrive every few weeks, and a rule written in August can be out of date by October. As a user I find this exciting, and the AI Integration page covers that side. As a teacher, I learned quickly how little rules and restrictions can do on their own, so I went back to the learning itself.

Students already know a good deal about how learning works. Without any training in education, they know that learning takes hours of practice. What they need is to connect that knowledge to the way they use AI. On the first day of lab I give them a one-page handout with three conditions for learning: cognitive friction, active recall over the illusion of competence, and productive failure. On the back, each student lists three ways AI helps learning and three ways it hurts, then swaps with a partner and tests the partner's points against the three conditions. In the first section that did this, the pooled answers arrived at all three conditions on their own, and the class AI policy was written from what students wrote.

A standard built this way stays usable when the tools change, because what helps people learn has changed very little.

Tracking Change Over Time

To see whether any of this is working, I document much of what I do, from the notes behind a redesigned quiz to the records of where students get stuck. I also keep trying new things, especially ones that use what I am good at.

Doing that for a year has changed how I read my own opening model. A statistical picture describes a system under one set of conditions at one moment. My teaching might be better described as a longitudinal study, where the same variables are watched as they shift from one semester to the next. Read that way, each semester is worth recording carefully, because the patterns I care about only show up across several of them.