Mixing Active Learning and Lecturing: Using Interactive Visualization as a Teaching Tool

class: center, middle, inverse, title-slide

# Mixing Active Learning and Lecturing: Using Interactive Visualization as a Teaching Tool
## JSM 2018
<div style='float:left'>
<hr color='#EB811B' size=1px width=720px>
</div>
 
### Jessica Minnier, PhD & Ted Laderas, PhD Oregon Health & Science University @datapointier
### July 29, 2018 Slides available at <a href="http://bit.ly/jsm-minnier" class="uri">http://bit.ly/jsm-minnier</a>

---

# Setting

## OHSU Data Science Institute

- 2 Day workshop
- 3 Hours for "Introduction to Statistics and Data Exploration"
- Aim of DSI: "bring together researchers, librarians, and information specialists for formal training on key topics in data science"

## Audience

- Librarians, information scientists, researchers
- Very little mathematical/programming background
- Heterogeneous background in science and research

---

# Goals

## Statistical Concepts

- Start with the didactics
- Use interactive visualizations to illustrate statistical concepts

## Data Exploration

- Empower students to explore data (no fear!)
- Encourage understanding of relationships of data

## Interactivity

- Interactive plots for exploration of multi-variable relationships
- Include some coding exercises (as bonus material)

---

# Methods

## Approach

- Implement as a LearnR Tutorial, but used with didactic teaching
- LearnR: uses Shiny to build interactive R Markdown style workbooks
- Can be deployed as a website, or on student’s computer (requires R/Rstudio)

## Practicalities

- Categorical data session and continuous data session
- Hosted on github as a package on Github laderast/DSIExplore (https://github.com/laderast/dsiexplore)
- Hosted workbooks on shinyapps.io for real time interactivity

---

# Interactivity

- Didactic lessons embedded in workbooks with interactive components
- Interactive sliders, dropdown options allow interaction with data filtering and analysis
- Interactive code teaches effect of changing code components on visualizations/analyses

---
class: clear
.font70[https://tladeras.shinyapps.io/categoricalData/]

![](ss1_learningobj.png)

---
class: clear

![](ss2_simpsonsparadox.png)

---
class:clear

![](ss3_chisquare.png)

---
class: clear

.font70[https://minnier.shinyapps.io/ODSI_continuousData/]

![](ss4_correlation.png)

---
class: clear

![](ss5_correlation2.png)

---
class: clear

![](ss6_correlation3.png)

---
class: clear

![](ss7_coding.png)

---
class: clear

![](ss8_ttest_sample.png)
---

# Conclusions & Results

- LearnR package + Shiny in R → interactive workbooks
- Students were empowered to learn
- Students liked the visualizations
 + “Very well done and methodical treatment - the sliders were great!”
- Students felt engaged with the subject
 + “Explanation of key statistical concepts was effective and really made me want to learn more.”
- Pre/Post-workshop survey: 95% of learners (survey responders) felt they gained practical knowledge (n=22)

---

# Survey Results

![](minnier_jsm2018_files/figure-html/unnamed-chunk-1-1.png)![](minnier_jsm2018_files/figure-html/unnamed-chunk-1-2.png)

---

# Impact

## Pros:

- Accessible to beginners
- Mathematical concepts are more memorable
- Sparks discussions
- Empowers and engages students in scientific discovery/analysis

## Cons:

- Advanced students may require more challenging activities
- Visualizations must be tested for effectiveness
- Requires programming skills to implement

---

# Future Work and Adaptations

- Expand materials with more advanced statistical concepts
- Longer workshops `\(\longrightarrow\)` more interactive material, more topics
- Determine which interactive explorations are most effective

## Introduction to Visualization/Data Literacy

- Extension of this work: https://tladeras.shinyapps.io/dataLiteracy/
- [HMSP410, Health Informatics](https://laderast.github.io/HSMP410/) for OHSU-PSU School of Public Health (co-taught by Ted Laderas and Bill Hersh)

---

# Further Information

- eCOTS e-poster: https://www.causeweb.org/cause/ecots/ecots18/posters/3-03
- Categorical Data: https://tladeras.shinyapps.io/categoricalData/
- Continuous Data: https://minnier.shinyapps.io/ODSI_continuousData/ 
- LearnR package: https://rstudio.github.io/learnr/ 
- DSIexplore LearnR package: https://github.com/laderast/DSIExplore

---

# Thank you!

Ted Laderas, PhD [laderas](https://twitter.com/laderas), [laderast](https://github.com/laderast/) https://laderast.github.io/

Contact me: minnier-[at]-ohsu.edu, [datapointier](https://twitter.com/datapointier), [jminnier](https://github.com/jminnier/)

Slides available at <http://bit.ly/jsm-minnier>

Code for slides available at <https://github.com/jminnier/talks_etc>

Slides created via the R package [xaringan](https://github.com/yihui/xaringan) by [Yihui Xie](https://twitter.com/xieyihui?lang=en) with the metropolis theme