Would love to be involved

[Aariq]

I'm teaching a mandatory recitation for an undergraduate biostatistics course next semester. This will be my third time teaching it. The first year, we used SPSS like in previous years. Then, I was able to convince the instructor to switch to R if I would stick around to teach the recitation. The first year using R went really well, I think, and we (me and another TA) learned a lot. This year I'm redesigning a bit to focus pretty exclusively on the Tidyverse. The only real hesitation before was that ggplot2 was more than most students needed for this course. But the different syntaxes used by different base R plots confused students, so I'm hoping it's worth the upfront investment into ggplot2, even if it seems like overkill, to make things less confusing to students by the end. I've never really taken a course that taught R, or really had any formal training about teaching R or any other programing language, so it's difficult for me to know how I'm doing except by getting student feedback. I'm hoping to get and give feedback on teaching materials and strategies!

[apreshill]

I have SO many thoughts for you. Firstly, this blog post:
http://varianceexplained.org/r/teach_ggplot2_to_beginners/

I have strong opinions that mirror David Robinson's- I haven't tried both ways, but I teach ggplot2 because I know and use ggplot2. I am currently teaching an intro biostats course where I relied on DataCamp's "in the classroom" program (re: you get free access for your students for the semester/quarter to all premium content!) to teach the tidyverse to start. Here is my course page:

https://ohsu-conj620.netlify.com

I also teach a dataviz course using ggplot2, materials are here: https://apreshill.github.io/data-vis-labs-2018/

Finally, I really recommend using ModernDive as a text (at least to start!). The first 5 chapters will get everyone up to speed quickly. The authors have also added sections to tell you which DataCamp courses complement the text.

I also really like the RStudio Data Science in a Box course for inspiration: https://github.com/rstudio-education/datascience-box

I found this post on the RStudio community site helpful too: https://community.rstudio.com/t/looking-for-best-ways-in-teaching-r-to-absolute-beginners/6998/3?u=apreshill

Hope that helps for now, and looking forward to continuing the conversation!

Alison

PS I have never had formal training on using R either 😄

[Aariq]

Thank you for all the wonderful links! I will check them out this weekend.

So I also read David Robinson's post about teaching ggplot2 to beginners, and I definitely think its a good idea. Last year I co-taught with someone who is a R wizzard, but whose brain works veeerrry differently than mine, and who finds base R to be more clear than Tidyverse. The switching back and forth was tough for students I'm sure, so this year it's going to be pretty exclusively Tidyverse, and I've convinced my new co-TA to drink the Tidyverse kool-aid. We're going to assign the Intro to R datacamp lesson, and maybe also the intro to the Tidyverse one (?) before the first recitation. We're also moving data visualization to the very first day, because I think it will help draw people in and see how much they can do with just a little coding.

I'm hesitant to assign more reading or textbooks since they already have the text for the lecture portion of the course to worry about, but I was planning on recommending ModernDive and r4ds as companions. Part of my hesitation is also that I think I will get some push back from the instructor in regards to ModernDive because she really wants to keep all example data biological. Perhaps some day I'll fork the ModernDive book and swap out all the datasets for ecology data...