Saturday, 22 June 2019

Annual Report to Stakeholders 2018-19


Every year in grad school I had to write a report on my research and academic progress. I found it a useful exercise so I've continued to do so as a faculty member and post-doc.


Summary: 



Professionally, this year was a struggle just to keep my head above water. I expect next year to be more productive in general, as well as more research oriented.





Gabriela makes me healthier and more productive in aggregate, but this year an unusually large part of my energy went into making a life together. Hense the net loss in research outout. This isn't a complaint, just an explanation that the lack of research this year is a temporary thing.


Personal: 


Gabriela and I got married in Canada in July, married again in Brazil in September, and recently married AGAIN at the Brazilian consulate in Canada for paperwork reasons. Married three times and neither a bigamist nor divorced, so there.

We also filed a PR application, which I'm personally counting as a publication because it took more research, care, and administration than most of the journal articles I've contributed to. The application totalled more than 100 pages, easily.


Teaching:



Four repeats (300, 300, 302, and 342), and two new classes (180 and 403) this year.
This courseload means that I've now taught 11 different classes in the span of 3 years. A department record, I suspect.

Stat 180 is a seminar course on careers in statistics. The bulk of the course is guest speakers from industry, government, and academia that are willing to talk to a group of undergrads about their jobs. This was a challenging course for me because it requires having a large enough network to be able to bring people in reliably.  I had to call in a lot of favours, and Dave Campbell in particular was extremely helpful in keeping this class afloat. Two posts, How to Give a Career Talk, with Question Prompts! and  Two career failure stories came out of this class.

Stat 403 was challenging for its material and its heterogenous student base. The title, at least for the undergraduate listing is "Intermediate Sampling". The actual course includes only a couple of weeks of sampling, while the rest of the course is based on design of experiments, mixed-effect, and split-plot models. I taught this class much as I had Stat 440 in the previous year – using a focus on guided projects to make up for theoretical depth in the material. Like Stat 440, it was a marginal success, but I would want to do some major restructuring, and more preparation, before I tackled something like this again.

Finding appropriate datasets was especially difficult, and in the second half of the course, I resigned myself to making synthetic datasets that had the features I wanted to use and highlight. This is where Dataset - The Giant Marmots of Moscow comes from.

The course is crosslisted with two graduate level service classes, so it included stats minors as well as graduate students from biology and physiology. All three of these groups had different strengths and were taking the course for different reasons. 

I had three lecture hours per week for both the grads and undergrads, a one-hour TA-led lab for the undergrads, and a one-hour seminar led by me for the grads. I tried to tailor these one-hour sessions to each group by giving the TA some R-programming material to teach, while I spent the grad student seminars on some of the more research-oriented facets of the material. Decent idea in hindsight, but execution still needs work.

I also assigned additional readings to the graduate students, two of which you can find in Reading Assignments - Split-Plot Design, Magnitude Based Inference. (Yes I am aware of the arbitrary nature of MBI, I still find it useful for showing how there are different ways to think about hypothesis tests)


Service:



- Three papers and a book were peer-reviewed. (Up from one paper and one book)

- 17 more papers were copy-edited for the Canadian Journal of Statistics up from 14 last year.

- 2 statistical replication reports were written for Meta Psychology.

- An extensive internal report on longitudinal student outcomes was written and presented to the department faculty. 


Publishing:



The Top Hat marketplace version of the book was a flop, and that's on me. The material I had prepared for Stat 300 wasn't quite ready for paid public consumption, and Top Hat works much better for large classes, whereas Stat 300 works much better for small classes.

I'm currently working on a traditional textbook format for the material, which itself has been greatly refined by two more semesters of iteration.

This year I also posted 31 new blog post and reworked 4 more, up from 24 new posts last year. Also, my R-centric posts now show up on the aggregator site R-bloggers!



Learning / Game Design:



Not much educational reading, but a lot of practical experience related to above. You can't have a personal life like the above without sacrificing something.

I did squeeze in time for some Jeopardy study, which also included a book of the New York Times' 'easy' crossword puzzles. I failed the Jeopardy exam for a third time, but at least I've gotten a lot better at crosswords.

I also started listening to some podcasts. For fun, it's been Hello Internet and Fill Me In, for education it's been A Lot to Learn, Stuff You Should Know, 99 Percent Invisible, and Data Bytes. By no means do I keep up with all of these, it has taken me half a year just to catch up on Hello Internet.





No comments:

Post a Comment