10 simple rules for teaching R for Data Science

class: top, left, title-slide

.title[
# 10 simple rules for teaching R for Data Science
]
.author[
### Tiffany Timbers (UBC)
]
.author[
### 2022-09-17
]
.date[
### bit.ly/timbers-cascadia-r-conf-2022
]

---

### 1. Teach R by doing data analysis

**What?** In your first lesson, load the data, do some "mild" wrangling and make a plot!

**Why?** Motivation!

***

.pull-left[
.tiny[
```r
library(tidyverse)

# load the data set
can_lang <- read_csv("data/can_lang.csv")

# obtain the 10 most common Aboriginal languages
aboriginal_lang <- filter(can_lang, 
                          category == "Aboriginal languages")
arranged_lang <- arrange(aboriginal_lang, 
                         by = desc(mother_tongue))
ten_lang <- slice(arranged_lang, 1:10)

# create the visualization
ggplot(ten_lang, aes(x = mother_tongue,
                     y = reorder(language, mother_tongue))) +
  geom_bar(stat = "identity") +
  xlab("Mother Tongue (Number of Canadian Residents)") +
  ylab("Language") 
```
]
]

.pull-right[
<img src="img/nachos-to-cheesecake-1.png" width="100%" style="display: block; margin: auto auto auto 0;" />
]

Source: 
[Chapter 1 of *Data Science: A First Introduction*](https://datasciencebook.ca/intro.html) 
by Tiffany Timbers, Trevor Campbell & Melissa Lee

---

.pull-left[
<blockquote class="twitter-tweet" data-lang="en"><p lang="en" dir="ltr">When you start writing a loop then turn it into dplyr <a href="https://twitter.com/hashtag/rstats?src=hash&amp;ref_src=twsrc%5Etfw">#rstats</a>; David Robinson (@drob) <a href="https://twitter.com/drob/status/701790246111825925">Feb 22, 2016</a></blockquote>
<script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
]

.pull-right[
<img src="img/jenny-teach-tidyverse.png" width="1736" style="display: block; margin: auto;" />

#### Explore more:
David Robinson:
- [Teach the tidyverse to beginners](http://varianceexplained.org/r/teach-tidyverse/)
blog post
- [Announcing "Introduction to the Tidyverse", my new DataCamp course](http://varianceexplained.org/r/intro-tidyverse/) blog post
- [Introduction to the Tidyverse](https://www.datacamp.com/courses/introduction-to-the-tidyverse) online course
]
---

### 2. Use participatory live coding!

**What?** Narrate and type out the code you are teaching. Have participants follow along.

**Why?** Demonstrates process and makes you appear human!

***

Source: [A video introduction to live coding part 2](https://www.youtube.com/watch?v=SkPmwe_WjeY&t=111s) by Lex Nederbragt

---

#### Explore more:
- [Ten quick tips for teaching with participatory
live coding](https://doi.org/10.1371/journal.pcbi.1008090) by Lex Nederbragt, Rayna Harris, Alison Hill & Greg Wilson
- [Live coding section of Teaching Tech Together](https://teachtogether.tech/en/index.html#s:performance-live) by Greg Wilson
- Example live coding videos by Lex Nederbragt ([bad example](https://www.youtube.com/watch?v=bXxBeNkKmJE&t=5s), [good example](https://www.youtube.com/watch?v=SkPmwe_WjeY&t=111s))
---

### 3. Give tons and tons of practice!

**What?** Give the learners many many problems to solve.

**Why?** Repetition leads to learning!

***

Read the data files listed in the table below and and bind the names listed in the table to them.

| File  | Object name to bind to the data frame | File location |
|---|---|----|
| `abbotsford_lang.xlsx`  | `abbotsford` | `data` directory of this repo |
| `calgary_lang.csv`  | `calgary`  | `data` directory of this repo |
| `edmonton_lang.xlsx`  | `edmonton`  | https://github.com/ttimbers/canlang/tree/master/inst/extdata |
|  `kelowna_lang.csv` | `kelowna`  | `data` directory of this repo |
| `vancouver_lang.csv`  | `vancouver`  | `data` directory of this repo |
| `victoria_lang.csv`  | `victoria`  | https://github.com/ttimbers/canlang/tree/master/inst/extdata |

---

.pull-left[
<img src="img/swirl.png" width="30%" style="display: block; margin: auto;" />

<img src="img/swirl-demo.png" width="90%" style="display: block; margin: auto;" />
]

.pull-right[
#### Explore more:
- [Swirl R package](https://swirlstats.com/) by Sean Kross and others
- [Practice worksheets to accompany Data Science: A First Introduction](https://github.com/UBC-DSCI/data-science-a-first-intro-worksheets#readme) by Tiffany Timbers, Trevor Campbell and Melissa Lee
- [Supervised Machine Learning Case Studies in R course](https://supervised-ml-course.netlify.app) by Julia Silge
- [Coding exercise types in Teaching Tech Together](https://teachtogether.tech/en/index.html#s:exercises) by Greg Wilson
]

---

### 4. Give tons and tons of timely feedback!

**What?** Automated software tests to give hints and feedback.

**Why?** Effective practice requires feedback!

***

---

.pull-left[
<img src="img/teaching-tech-together.jpeg" width="75%" style="display: block; margin: auto;" />
]

.pull-right[
#### Explore more:
- [Automatic Grading chapter of *Teaching Tech Together*](https://teachtogether.tech/en/index.html#s:exercises-grading) by Greg Wilson
- [learnr](https://rstudio.github.io/learnr/) R package for automated feedback of R Markdown documents
- [nbgrader](https://nbgrader.readthedocs.io/en/stable/) tool for automated feedback & autograding of Jupyter notebooks
- [ottergrader](https://otter-grader.readthedocs.io/en/latest/) tool for automated feedback & autograding of both Jupyter notebooks & R Markdown documents
]

---

## 5. Use tractable or toy data examples

**What?** Use a countable number of things when teaching a new tool, method or algorithm.

**Why?** Allows one to see how all the elements are manipulated.

***

A small subset of the `palmerpengiuns` data set for teaching K-means clustering:

Source: [Clustering chapter of *Data Science: A First Introduction*](https://datasciencebook.ca/clustering.html) by Tiffany Timbers, Trevor Campbell & Melissa Lee

---

.pull-left[

.small[
```r
left_join(superheroes, publishers)
```
]

We basically get `x = superheroes` back, but with the addition of variable `yr_founded`, which is unique to `y = publishers`. Hellboy, whose publisher does not appear in `y = publishers`, has an `NA` for `yr_founded`.

<img src="img/jenny-bryan-left-join.png" width="100%" style="display: block; margin: auto;" />
]

.pull-right[
#### Explore more:
- [Jenny Bryan's join cheat sheet](https://stat545.com/join-cheatsheet.html)
- [`dpylr::tibble`](https://tibble.tidyverse.org/reference/tibble.html) or [`dpylr::tribble`](https://tibble.tidyverse.org/reference/tribble.html) R functions
- [`sample`](https://www.rdocumentation.org/packages/base/versions/3.6.2/topics/sample) or [`dpylr::sample_n`](https://dplyr.tidyverse.org/reference/sample_n.html) R functions
- [charlatan](https://docs.ropensci.org/charlatan/) R package for creating fake data
- [drawdata.xyz](https://drawdata.xyz/) app to draw a data set and download it
]
---

## 6. Use real and rich, but accessible, data sets

**What?** Use real and interesting data that is also easily understood by all.

**Why?** Motivating!

***

Source: [`canlang`](https://ttimbers.github.io/canlang/) R package
---
.pull-left[
Jenny Bryan's [`gapminder`](https://github.com/jennybc/gapminder) R data package:

.pull-right[
<img src="10-simple-rules-for-teaching-r-for-data-science_files/figure-html/gapminder-1.png" style="display: block; margin: auto;" />
]

#### Explore more:
- [`canlang`](https://ttimbers.github.io/canlang/) R data package 
- [`palmerpenguins`](https://allisonhorst.github.io/palmerpenguins/) R data package
- [`Lahman`](https://github.com/cdalzell/Lahman) baseball R data package
- [`datateachr`](https://github.com/UBC-MDS/datateachr) R data package
- [`taxyvr`](https://github.com/UBC-MDS/taxyvr) R data package
- [UCI Machine Learning Repository](https://archive-beta.ics.uci.edu/) for data sets

---

## 7. Provide cultural and historical context

**What?** When R things seem "odd", take time to explain why they are the way they are.

**Why?** Cultivates an appreciation, or at least understanding of R's quirks (compared to other programming languages). Helps prevent frustration and/or annoyances. Helps with motivation again!

***

When teaching programming with the `tidyverse`, 
I make sure to introduce R's history and it's design purpose, 
so that the role and rationale for unquoted column names is clear:

>#### Now, a bit of history about R
>- An implementation of the S programming language (created at Bell labs in 1976)
>- written in C, Fortran, and R itself
>- R was created by Ross Ihaka and Robert Gentleman (Statisticians from NZ)
>- R is named partly after the authors and partly as a play on the name of S
>- First stable beta version in 2000

---

.pull-left[
<img src="img/why-assignment-1.png" width="100%" style="display: block; margin: auto;" />

(Or why `<-` instead of `=`?)

Source: https://colinfay.me/r-assignment/
]

.pull-right[

#### Explore more:
- [History and Overview of R chapter](https://bookdown.org/rdpeng/rprogdatascience/history-and-overview-of-r.html) from R Programming for Data Science by Roger Peng
- [Introduction to *R for Data Engineers*](https://tidynomicon.github.io/tidynomicon/) by Greg Wilson
- [Wikipedia article on R](https://en.wikipedia.org/wiki/R_(programming_language)
]
---

## 8. Build a safe, inclusive and welcoming community

**What?** Put in place scaffolding and guidelines to facilitate a safe learning environment.

**Why?** Students cannot effectively learn if they do not feel safe.

***

Source: https://ubc-dsci.github.io/dsci-100-student/CODE_OF_CONDUCT.html
---

Source: https://docs.carpentries.org/topic_folders/policies/code-of-conduct.html

#### Explore more:
- [Ally Skills workshop](https://frameshiftconsulting.com/ally-skills-workshop/)

---

### 9. Use checklists to focus and facilitate peer learning

**What?** Focus and provide structure to student peer review by providing them review checklists.

**Why?** Peer review is hard and often new to many. Also reviewing something you are new at is even harder! What to focus on?

***

##### Data analysis review checklist:

.pull-left[
<img src="img/da-review-checklist-1.png" width="85%" style="display: block; margin: auto;" />
]

.pull-right[
<img src="img/da-review-checklist-2.png" width="85%" style="display: block; margin: auto;" />
]

Source: https://github.com/UBC-MDS/data-analysis-review-checklist
---

#### ROpenSci R package reviews utilize checklists:

.pull-left[
<img src="img/ropensci-1.png" width="80%" style="display: block; margin: auto;" />
]

.pull-right[
<img src="img/ropensci-2.png" width="80%" style="display: block; margin: auto;" />
]

Source: https://github.com/ropensci/software-review/issues/495

#### Explore more:
- [Using rOpenSci Software Peer Review Guidelines for Teaching](https://ropensci.org/blog/2019/08/27/software-peer-review-guidelines-for-teaching/) blog post by Tiffany Timbers
- [ROpenSci review template](https://devguide.ropensci.org/reviewtemplate.html)
- [Journal of Open Source Software review checklist](https://joss.readthedocs.io/en/latest/review_checklist.html)
- [Machine Learning Reproducibility Checklist](https://www.cs.mcgill.ca/~jpineau/ReproducibilityChecklist.pdf) by Joelle Pineau
- [ Checklist Manifesto](http://atulgawande.com/book/the-checklist-manifesto/) by Atul Gawande

---

## 10. Have students do projects

**What?** Have students perform a (scoped) data analysis project on a topic of their choosing.

**Why?** Motivation and exposure to the messiness of real data analysis.

***

Student project from DSCI 524 (Collborative Software Development):

Source: https://github.com/UBC-MDS/rlyrics

---

Projects are central to UBC's founding data science course, STAT 545 (created by Jenny Bryan).

Source: https://stat545.stat.ubc.ca/

#### Explore more:

- [Project courses in MDS](https://ubc-mds.github.io/2019-08-22-project-courses/) blog post by Tiffany Timbers and Mike Gelbart
- [Project-based learning: A review of the literature](https://doi.org/10.1177/1365480216659733) by Dimitra Kokotsaki, Victoria Menzies and Andy Wiggins

---

## 10 simple rules for teaching R for Data Science

1. Teach R by doing data analysis

2. Use participatory live coding

3. Give tons and tons of practice

4. Give tons and tons of feedback

5. Use tractable or toy data examples

6. Use real and rich, but accessible, data sets

7. Provide cultural and historical context

8. Build a safe, inclusive and welcoming community

9. Use checklists to focus and facilitate peer learning

10. Have students do projects

---

### Acknowledgements

- UBC Master of Data Science students

- UBC DSCI 100 and 310 students

- [UBC Master of Data Science teaching team](https://ubc-mds.github.io/team/) and the DSCI 100 teaching team

- [The Carpentries](https://carpentries.org/)

- [Greg Wilson](https://third-bit.com/), Software Engineer, Deep Genomics

- [Jenny Bryan](https://jennybryan.org/), Software Engineer, Posit

- [David Robinson](http://varianceexplained.org/), Director of Data Scientist, Heap

- [Roger Peng](https://rdpeng.org/), Professor of Statistics and Data Sciences , University of Texas, Austin

- [Lex Nederbragt](https://lexnederbragt.com/), Senior Lecturer, University of Oslo

- [Sean Kross](https://seankross.com/), Data Staff Scientist, Fred Hutch Data Science Lab

- [Colin Fay](https://colinfay.me/), Data Scientist & R Hacker, ThinkR

- And many more educators in the R community!!! There are too many to list!

---

class: inverse, center, middle

## 10 simple rules for teaching R for Data Science

#### talk slides: *[https://bit.ly/timbers-cascadia-r-conf-2022](https://bit.ly/timbers-cascadia-r-conf-2022)*