a course I designed and have not yet taught
Systems Engineering in Public Programs · a graduate course designed backward from its big ideas · syllabus, course map, grading plan, and assignment design
A graduate course on systems engineering in public programs, designed backward from its big ideas. Students build models for public-interest decisions where the stakeholders disagree about what counts as better. I have not taught it. I am publishing it as a course proposal, because it is the clearest statement I can make of what I would build if a department gave me a slot.
- Syllabus
- Course map, fifteen weeks session by session
- Course foundations planner, the backward design
- Grading plan
- Transparent assignment design
What it is built around
The transfer goal is that students can design a decision model for a social setting and then evaluate what their own modeling choices did. The essential questions are the ones I want them arguing about in week fifteen as much as in week one. What is a good model, and good for whom? What is the difference between modeling a technical problem and modeling a social one? Is a good model for one problem a good model for another?
The content moves through multiple objective decision making, data envelopment analysis, public health preparedness, network interdiction and attacker-defender models, and multicriteria ranking. Each block pairs a methods session with a case study session, and the case is where the stakeholders appear.
How it is assessed
The grading plan is deliberately spread across many small things rather than two large ones. A pre-assessment quiz, an initial feedback survey that I use to adjust the course, discussion posts, in-class discussion, participation in activities, a game, peer and instructor tutorial sessions, at-home exercises in Jupyter notebooks, a simulation, and a three-part project with peer evaluation at every presentation.
The pre-assessment and the feedback survey are there for the same reason as everything else on this site. I cannot know what class to teach until I know who is in the room.
Auditing someone else’s course against AI
In Summer I 2026 I revised the course materials for IE 6200, Engineering Probability and Statistics, a graduate course taught by Prof. Doug Schmucker. It was a materials appointment rather than a taught section.
I reviewed the material from a student perspective and analyzed the assignments against AI usage, creating answers with different prompts to test how robust each assignment was to the use of artificial intelligence. The finding that mattered was not that students can use a model to answer a question. It is that the robust assessments were the ones asking for something the model cannot supply: the student’s own data, the student’s own context, or a defence of a choice against the alternatives they rejected. I met with the instructor weekly and delivered a written report on the full set of materials.
That audit is now part of how I design assessment, and it is why the project rubric on the exercises page puts 25 points on justifying a modeling choice rather than on producing the model.
The assignment design worksheet is the piece I would point a hiring committee to. It states, for every assignment, the purpose, the skills being practiced, what students will actually do, and the criteria for success, in language written for students rather than for a syllabus.