I only understand a problem when I find a way to teach it.

My teaching focuses on methodology for causal inference. I teach researchers who generate or repurpose data to support decision-making, especially for clinical medicine and health policy. My aim is to help them develop critical thinking, problem-solving skills, and the ability to articulate sound arguments.

I have developed and led core methods courses at the Harvard School of Public Health and Harvard Medical School. I currently teach clinical epidemiology at the Harvard-MIT Division of Health Sciences and Technology. Also, HarvardX offers my free online course Causal Diagrams: Draw Your assumptions Before Your Conclusions.

If you want to come to study at Harvard, or to take a course outside Harvard, my colIeagues and I offer the following:

As far as I remember, I’ve always wanted to be a teacher.

  • Why I teach

    Teaching arises from my need to understand. I would prepare a teaching plan even if I were far from classrooms, because a teaching plan helps me understand a subject. I can imagine Beethoven writing a string quartet even without any hope of hearing it performed (not that I’m comparing myself to Beethoven).

    Engaging with students in a discussion that I’ve framed—and they often reframe—is both invigorating and humbling. Learning becomes a two-way road. I may help students grow, perhaps even shape how they see the world, but I also encounter perspectives that prompt me to reconsider key ideas and assumptions.

    What begins as a private urge to understand becomes a collective enterprise: one I guide, but do not rule.

  • What I teach

    I teach methods for drawing causal inferences from data. In my courses, we examine what conditions must hold for causal claims to be warranted and the tools, from study design to statistical analysis, that can support causal inference when those conditions are met.

    While my examples often come from medicine and public health, the underlying methods are relevant to any field that requires causal inference, including economics, sociology, law, public policy…

    My aim is to help students both evaluate existing evidence and produce new evidence that can inform real decisions—by patients, clinicians, regulators, and the public. My courses emphasize critical thinking, problem-solving skills, and the ability to articulate sound arguments.

  • How I teach

    I try to decompose complex issues into simple steps. Teaching is then like climbing a mountain with students: each step is manageable, and by the end they can enjoy the view from the top. Along the way, frequent reminders of the destination—plus a bit of humor and suspense in the storytelling—help keep us moving together toward that final payoff.

    Lectures are, in my view, an efficient way to present a new topic; subsequent small-group deliberations deepen and refine understanding. In data analysis courses with a hands-on approach, students bring their computers for practical exercises.

    In longer courses, we take on complex real-world problems. Students develop and defend their views using the available data. When disagreements emerge, I help students sharpen their arguments and make explicit the assumptions that lead to different interpretations of the evidence.