Featured post

Textbook: Writing for Statistics and Data Science

If you are looking for my textbook Writing for Statistics and Data Science here it is for free in the Open Educational Resource Commons. Wri...

Tuesday 26 December 2023

Reflection on an undergrad/grad crosslisted design of experiments course


I'm looking to re-examine and improve my teaching since I started working as a professor again for the University of Waterloo, and part of that means writing reflections on courses again. This one is for a 4th year / graduate level crosslisted course on design and analysis of experiments.

What worked:

- Asking for feedback around week 4-5. Students felt heard. They had some concerns that I was able to address. Namely, that I was teaching in a data science style when I should have been focusing more on mathematical rigour.

Also, I was able to pivot my lecturing style to from typing in ms paint to writing on paper, which seemed to work. Over the holidays, I bought an artist's tablet and I'm learning to use it effectively, so soon I'll be able to eliminate the paper and just upload handwritten annotations directly.

- Digital textbook. Montgomery is the go-to book for this course, and thankfully an older version of the book was available digitally through the library. I managed to rely on a textbook while saving the class a lot of money.

- Recording class. This continues to be well received. Although I need to find a recording system that works without the internet. Currently I'm using Zoom.

What didn't:

- Leaving semester prep to the last minute. This is especially true because Design of Experiments has always challenged me - factorial designs aren't something I use in practice much, so I was reliant on the text book and Nathaniel Steven's notes a lot. In my defense, I was taking care of an 8-month old without any day care available yet, but that challenge will always be there in some form and I need to learn to work around it.

- Pacing. I ran out of planned material with 3 lectures to go, so we did a lot of final exam review. This would have been better if it was spread throughout the semester as additional examples.

- Treating 430 and 830 as a the same class. I wonder if there are any resources for crosslisted classes. I wonder if I should write one. Would make a good M.Ed. thesis topic.

Challenges (excuses and whining):

- 1.5 new preps this term. (Stat 230 isn't really a new prep because it's so familiar)

- Baby

- Constantly sick

Challenges (actionable)

- Lack of deep knowledge about Stat 430

- Balancing exam difficulty. It's hard to come up with questions that some people will get and others won't in a way that's fair. This comes back to the lack of expertise. I think I skewed on the easy side by looking at the results. Waterloo students are VERY smart as well, so an exam that would work elsewhere tends to skew easy in these classes. It's also not a huge problem in a 4xx/8xx class to have a 78% midterm average because the 8xx students have substantially higher grade requirements anyways.

- Classroom engagement. This might be an effect of the student body involved, but I had very little direct engagement during class. I had quite a bit of student activity on piazza, and I must remember to be more communicative on there - it's not enough just to address the problems people ask on piazza, they need to see the response right there on piazza that something was done.

- Participating in Marking. Still on the fence about this. This class has a lot of written answers on tests and assignments, and I've taken to marking a lot of those myself instead of relying on TAs. It takes time, some of which is offset by reduced remark requests. It also gives TAs less experience with marking written questions, and I should remember that I'm also in the business of training them.

No comments:

Post a Comment