Self-Assessment with learnr

class: center, middle, inverse, title-slide

# Self-Assessment with <code>learnr</code>
## useR Boston Meetup
### Maria Kamenetsky
### University of Wisconsin-Madison
### February 24, 2021

---

#Overview

- Background & Motivation
- `learnr` Basics
- Sharing Your Tutorial
- Demo
- Wrap-Up

---

# Background

.center[![description of the image](rstudioconf.png)]

rstudio::global(2021) Diversity Scholarship
  - Mine Çetinkaya-Rundel workshop: **“What they forgot to teach you about teaching R”**
  - Alison Hill: “Introducing Yourself Online”

---

# The package: `learnr`

The **learnr** package can be used to turn *R Markdown* documents into interactive tutorials.

Can easily incorporate figures, equations, videos, interactive *Shiny* components. 
For more information see https://rstudio.github.io/learnr/

--
**Goal**: Use `learnr` for student *self-assessment*

---

# Motivation

STAT679-III: Spatial Statistics for Lattice Data

Co-developed with Dr. Jun Zhu

Focus on lattice data, exploratory spatial data analysis (`tidyverse`), Moran's I & Geary's C (`spdep`), simultaneous autoregressive (SAR) and conditional autoregressive models (CAR) (`spatialreg`)

Cognitive overhead for learners to learn new material *and* review previous statistics materials and `R`.

---

# The Audience

Professional MS in [Environmental Observation & Informatics](https://nelson.wisc.edu/graduate/professional-programs/environmental-observation-informatics/index.php)
  - Diverse background (engineering, ecology, GIS, ...)

Course sequence:
  - Summer: 
    - Statistics: Foundations and Regression With Big Data
    
  - Spring Semester:
    - **5 week course on Spatial Statistics for Lattice Data **

---

# Self-Assessment Tool

https://github.com/mkamenet3/stat679selfassessment

.center[![description of the image](satool.png)]

---
# `learnr` Basics

After installing the `learnr` package:

- `File` -> `New File` -> `R Markdown` -> (From Template) Interactive Tutorial

```
---
title: "STAT679-III: Self-Assessment"
output: learnr::tutorial
runtime: shiny_prerendered
---
```

---

# `learnr` Basics

````markdown
```{r setup}
data("mtcars")
mod <- lm(mpg ~ wt, data=mtcars)
par(mfrow=c(1,2))
plot(mod, which=c(1,2))

```
````

````markdown
*```{r myquiz}
library(learnr)
quiz(
  question("In the above diagnostic plots,
  the plot on the *left* is:",
           answer("Residuals vs. fitted plot", correct=TRUE),
           answer("QQ plot"),
           answer("I'm not sure"),
           allow_retry = TRUE,
           try_again_button = "Try again",
           correct="Correct!")
           )
```
````

---

# `learnr` Basics

````markdown
```{r setup}
data("mtcars")
mod <- lm(mpg ~ wt, data=mtcars)
par(mfrow=c(1,2))
plot(mod, which=c(1,2))

```
````

````markdown
```{r myquiz}
*library(learnr)
quiz(
  question("In the above diagnostic plots,
  the plot on the *left* is:",
           answer("Residuals vs. fitted plot", correct=TRUE),
           answer("QQ plot"),
           answer("I'm not sure"),
           allow_retry = TRUE,
           try_again_button = "Try again",
           correct="Correct!")
           )
```
````

---

# `learnr` Basics

````markdown
```{r setup}
data("mtcars")
mod <- lm(mpg ~ wt, data=mtcars)
par(mfrow=c(1,2))
plot(mod, which=c(1,2))

```
````

````markdown
```{r myquiz}
library(learnr)
*quiz(
  question("In the above diagnostic plots,
  the plot on the *left* is:",
           answer("Residuals vs. fitted plot", correct=TRUE),
           answer("QQ plot"),
           answer("I'm not sure"),
           allow_retry = TRUE,
           try_again_button = "Try again",
           correct="Correct!")
           )
```
````

---

# `learnr` Basics

````markdown
```{r setup}
data("mtcars")
mod <- lm(mpg ~ wt, data=mtcars)
par(mfrow=c(1,2))
plot(mod, which=c(1,2))

```
````

````markdown
```{r myquiz}
library(learnr)
quiz(
*  question("In the above diagnostic plots, 
*  the plot on the *left* is:", 
           answer("Residuals vs. fitted plot", correct=TRUE),
           answer("QQ plot"),
           answer("I'm not sure"),
           allow_retry = TRUE,
           try_again_button = "Try again",
           correct="Correct!")
           )
```
````

---

# `learnr` Basics

````markdown
```{r setup}
data("mtcars")
mod <- lm(mpg ~ wt, data=mtcars)
par(mfrow=c(1,2))
plot(mod, which=c(1,2))

```
````

````markdown
```{r myquiz}
library(learnr)
quiz(
  question("In the above diagnostic plots,
  the plot on the *left* is:",
*           answer("Residuals vs. fitted plot", correct=TRUE), 
*           answer("QQ plot"), 
*           answer("I'm not sure"), 
           allow_retry = TRUE,
           try_again_button = "Try again",
           correct="Correct!")
           )
```
````

---

# `learnr` Basics

````markdown
```{r setup}
data("mtcars")
mod <- lm(mpg ~ wt, data=mtcars)
par(mfrow=c(1,2))
plot(mod, which=c(1,2))

```
````

````markdown
```{r myquiz}
library(learnr)
quiz(
  question("In the above diagnostic plots,
  the plot on the *left* is:",
           answer("Residuals vs. fitted plot", correct=TRUE),
           answer("QQ plot"),
           answer("I'm not sure"),
*           allow_retry = TRUE,
*           try_again_button = "Try again",
*           correct="Correct!")
           )
```
````

---

# Sharing Your Tutorial: (1) GitHub

Share your `learnr` tutorial as a package:

```
*#install from GitHub
*install.packages("remotes")
*remotes::install_github("mkamenet3/stat679selfassessment")

#Run tutorial locally
library(stat679selfassessment)
learnr::run_tutorial("selfassessment", package = "stat679selfassessment")
```
GitHub: https://github.com/mkamenet3/stat679selfassessment

---

# Sharing Your Tutorial: (1) GitHub

Share your `learnr` tutorial as a package:

```
#install from GitHub
install.packages("remotes")
remotes::install_github("mkamenet3/stat679selfassessment")

*#Run tutorial locally
*library(stat679selfassessment)
*learnr::run_tutorial("selfassessment", package = "stat679selfassessment")
```

GitHub: https://github.com/mkamenet3/stat679selfassessment

---

# Sharing Your Tutorial: (2) Use mybinder.org

Create 3 additional files in your directory:

--

1) `install.R`

Define the packages to be installed:
```
*install.packages("learnr")
*install.packages("MASS")
*install.packages("knitr")
```
--

2) `runtime.txt`

Define the `R` version to use in the Docker image and the date:

```
*r-3.6-2021-02-12
```
--

3) `postBuild`

Move the tutorial R Markdown file to the working directory so it launches from mybinder.org:
```
*mv inst/tutorials/selfassessment/ .
```

---
# Demo

---
# Acknowledgements

- RStudio
- Mine Çetinkaya-Rundel 
- `learnr` [developers](https://cran.r-project.org/web/packages/learnr/index.html) (Barret Schloerke, JJ Allaire, Barbara Borges, Angela Li, RStudio, Ajax.org B.V., Zeno Rocha, Nick Payne, Jake Archibald)
- Ted Laderas, [*Getting LearnR tutorials to run on mybinder.org*](http://laderast.github.io/2020/09/15/getting-learnr-tutorials-to-run-on-mybinder-org/)

---

class: center, middle

# Thank you!

@mkamenets

https://mariakamenetsky.com/