Some notes about RStudioConf 2019.
Well, RStudio Conf 2019 has come and gone. I attended the main conference, starting with the poster session on wednesday and stayed through the tidyverse developer day on Saturday.
To say that the conference was inspiring was an understatement. So many talented people working on such interesting and inspiring packages! It made me excited again about doing data science and teaching data science.
This post is going to highlight the interesting talks about education and organizational management at the conference. There were lots of thought provoking talks and great resources about education. Here’s a link to the other resources, posters, and talks (thanks, Karl Broman!). There are a lot of new neat whiz bang features in RStudio, which I won’t cover, but are covered very well by others at Karl’s link above.
UPDATE: I’m adding links to the videos as subheadings that you can click on. I won’t embed these directly because there is often related material with the video.
There was much sharing of educational resources. I’ll talk about a few.
Carl Howe (Director of Education for RStudio) presented a number of resources that RStudio is providing for educational purposes. Their goal is to train the next million R users. Here are some of the RStudio based educational resources that might be useful for everyone.
RStudio Cloud is now available free for people who teach courses, who can send them a course syllabus to gain access. RStudio Cloud includes resources to make coursework publically or privately available, allows instructors to install default packages, and gives students an immediate way to start playing with R/RStudio. We are definitely going to use this in teaching our BMI 569/669 Data Analytics course. Mel Gregory’s talk was about the nuts and bolts of using RStudio Cloud in a classroom situation. Highly recommended to get familiar with the basics.
Mine Çetinkaya-Rundel also has a Data Science In a Box course available that can be forked and used by anyone. There is a great discussion of how to use the course and the tech stack (RStudio Cloud, GitHub, and Slack) needed to make the course runnable.
There are also a number of LearnR Tutorials built into RStudio Cloud called RStudio Primers that cover a lot of basic RStudio operations for the self-directed learners.
Jessica Minnier and I presented at the poster session about our LearnR tutorials, DSIExplore and dataLiteracyTutorial and our
burro package, which lets users explore a new dataset with a simplified Shiny App. There’s even a function that lets you build a shiny app from a dataset that you can publish to a Shiny server such as
shinyapps.io or RStudio Connect for sharing and having a data scavenger hunt together. Or as Angela Bassa calls it, an EDA Party. Whoo!
I really loved Irene Stevens’ talk about Teaching Data Science Using Puzzles. Irene and Jenny Bryan put together a simple project-based framework that lets students download simple data wrangling puzzles per week, lets them submit their answer to an answer server and get feedback, and when they solve it, it creates a minimal reproducible example (reprex) to share with their puzzle community. Very Slick!
Kelly Bodwin gave a presentation of how she used Shiny apps as an intermediate step as part of an approach to teaching introductory statistics. Kelly is of the opinion that coding is good even for non majors, and her Shiny apps provide an intermediate step between a more click-based app and coding, asking the students to input their variables, their hypothesis, and highlights the output that results. I really liked this approach. I also really loved this slide: