Teaching Philosophy
Dr. Rocío Mercado Oropeza
I teach machine learning to chemists and chemistry to computer scientists, both in my courses and in my research group. Most of my students are learning some of the material from outside their own discipline, so much of my job comes down to working out what a student already knows and building on it. In my teaching I focus on inclusion, on alignment between what I teach and what I assess, and on testing and revising my own approach.
To me, inclusion means supporting each student’s academic and personal growth, helping them build a scientific identity, and remembering that every student comes in with a different background. Students with a strong scientific identity are more likely to commit to a science career and to stay in STEM, and the effect is especially strong for students from groups historically excluded from science [1, 2]. The belief that success depends on innate “brilliance” works in the other direction. Fields that hold this belief have fewer women and fewer students from other underrepresented groups [3], and children pick up these stereotypes by about age six [4]. I try to teach a growth mindset. Being “good” or “bad” at math or programming changes with practice, and a mistake or a gap in preparation is not a measure of ability. Instructors’ beliefs also matter. In STEM courses taught by faculty who think ability is fixed, racial achievement gaps are about twice as large and students report lower motivation [5]. I therefore build inclusion into the design of my courses [6]. I use formats such as think-pair-share and low-threshold question channels, which give quieter students and second-language speakers a way to take part. I tell students plainly that difficulty and mistakes are part of learning, and I include work by scientists from underrepresented groups in the curriculum. A sense of belonging [7] also grows outside the lecture hall, in office hours, in the lab and at conferences. At conferences I make time to go to junior researchers’ talks and talk with them afterward, because a small moment like that can shape how a young researcher sees themselves as a scientist.
Students learn what the assessment asks of them, and that isn’t always what the lecturer emphasizes. Following Biggs’s idea of constructive alignment [8], I start from the intended learning outcomes and design the activities and assessments around them. In AI for Molecules, a course I designed from scratch, I built in alignment and scaffolding from the start. Each session opens with a short recap, often in pairs, and we then work through increasingly complex examples together to build up the concepts.
I want students to leave my courses able to use generative AI well, since the fields they are going into already rely on it. Restrictive policies are hard to enforce fairly, because AI-assisted work is difficult to detect [9], and they end up penalizing the students who follow the rules. I allow these tools openly and hold students responsible for everything they hand in, however they produced it. Short, low-stakes oral checks on a few randomly chosen submissions back this up. I see this as an alignment question as well. AI assistance can improve the finished work without improving the learning behind it [10, 11], and if a model can complete a task without the student doing the thinking, that task was never a good measure of learning. I am moving my assessments toward judgment, critique and explaining choices, and I state how much AI use is allowed for each task [12]. I expect to keep adjusting this as the tools change [13].
In research mentoring, I try to teach the critical thinking that research needs: finding an important problem, shaping it into a project, carrying it out and communicating the results. Most of the supervision problems I have seen came from expectations nobody had said out loud. I set clear expectations at the start of each mentoring relationship and come back to them over time. To keep things transparent, I maintain a regularly updated group guide and AI policy for my team, which set out how we work together, how we handle authorship and what we expect of each other, and we agree on authorship at the start of each project. Each person meets with me one-to-one every one to two weeks. Separately, I offer development meetings about their longer-term goals, whether in academia, in industry, or any other professional question they want to talk through. Many of our projects are co-supervised with industrial partners, so students also see what research looks like outside academia. I want my group to be a place where people can say an idea didn’t work, or that they are stuck, without worrying about being judged. I think of independence as something I hand over in stages the student can see. They first carry out a project, then make its methodological decisions, then supervise others, and eventually formulate the research question themselves. Teaching is part of this, and several of my doctoral students have developed their own teaching as assistants in my courses.
My nearly four years as an assistant professor have taught me that every student is an individual and that no single approach to mentoring works for everyone. What worked for me may not work for someone else, so I try to find out what motivates each person and adjust how I support them, including as they change over time. When I take someone into my group, I make a commitment to support them to the end of their degree or position. I see supervision and mentorship as a partnership and ask for the same commitment in return. If either of us is not keeping up our side, I hope we can bring it up early and directly, so that we can agree together on how to move forward.
I also treat my teaching as something to study. I form a hypothesis about what will help students learn and collect evidence from attendance, course evaluations, student representatives, teaching assistants and colleagues who sit in on my classes. When a change doesn’t have the effect I expected, I revise it. My teaching draws directly on my research. AI for Molecules grew out of the problems my group works on, so students learn current methods and engage with open questions. I share my course materials and what I learn with colleagues, and I have developed assessments together with other examiners. I keep developing as a teacher and mentor through formal training, including the Chalmers Diploma in Teaching and Learning in Higher Education, through reading the higher-education literature, and through feedback from colleagues, students and the people I mentor.
Last updated: October 16, 2026.