Data Storytelling

class: center, middle, inverse, title-slide

# Data Storytelling
## BioData Club
### Ted Laderas, PhD Division of Bioinformatics and Computational Biomedicine Department of Medical Informatics and Clinical Epidemiology
### 2019-11-15 <a href="https://twitter.com/tladeras">tladeras</a>

---

# Overview

- Part 1: Getting Started
- Part 2: Decluttering
- Part 3: Annotating
- Part 4: Highlighting
---
# Future Events

- December 10 (12 PM) - Data and Donuts
- January 10 (4:30 - 6 PM) - How to make a reproducible paper

Make sure you're signed up on the mailing list for more info! Link at http://biodata-club.github.io

---
# Reminder

This workshop adheres to the BioData Club Code of Conduct:

https://biodata-club.github.io/code_of_conduct

Please be respectful of your fellow learners and help each other learn.

Remember, it's dangerous to learn alone! So partner up with someone, it's fun to learn together.
---
# Motivation: Exploratory versus Explanatory

.pull-left[
**Exploratory analysis**: 
  - exploring and understanding the data, conducting the analysis
]
.pull-right[
**Explanatory analysis**: 
  - explaining your findings from your analysis in a coherent narrative that leads to a call to action
]
---
# Effective Visual Communication

Focus on three techniques:

- Decluttering your plot
- Annotating your graph and data
- Highlight data using Preattentive Attributes

---
# Paper Doll Approach

.pull-left[
- We're going to take a basic plot and dress it up
- Modify its appearance to make our point more understandable and immediate
]

.pull-right[
![](image/giphy.gif)
]
---

class: center, middle
# Let's open 01-starting-out.Rmd

---
class: center, middle
# Part 2: Declutter your graph

---
# Why do we need to declutter our graphs?

- Reduce cognitive load (help tired and cranky viewers)
- Viewer can focus on what matters
- Not all information is useful for your viewer

---
# Example: London Subway Diagram

.pull-left[
![](image/london-underground.jpg)
]

.pull-right[
- Triumph of minimal design
- Removes geography
- Emphasizes: what lines to I take to get from A to B?
]

.footnote[http://theconversation.com/sublime-design-the-london-underground-map-26240]

---
# Cognitive Load

- Think of your audience: 
    - Tired and cranky and want you to get to the point!
    - Short term memory 5 +/- 2 things at once
- Remove elements that distract from your message
    - Shadows, 3D Effects
    - Legends/too many colors 
    - Axis Titles (sometimes)

---
# Ask Yourself

- Does this element support [the point I want to make about the data?](http://www.storytellingwithdata.com/blog/2017/3/29/declutter-this-graph)

---
class: center, middle
# Let's open up 02-decluttering.Rmd

---
class: center, middle
# Part 3: Annotating Your Graphs

---
# Guiding Your Viewer

Another way we can guide people through our visualization is by using **annotations**, which can be very helpful in guiding someone through our figure. Let's review some best practices.

---
# Use your titles/captions!

- Titles can guide people to the point of your figure
- Primes people to know what to look for
- "If there is a conclusion you want your audience to reach, state it in words" - Cole Nussbaum Knaflic
---
# Don't label everything

.pull-left[- Think about only labeling the data that matters
- If you want someone to compare two shows, label them
- Think about groupings and "super categories" to help your viewers make sense of the graph]

.pull-right[
<img src="image/super-category.png" width="727" />
]
---
class: center, middle
<img src="image/Colin.png" width="2000" />

https://datawoj.co.uk/visualising-data-on-which-social-media-network-us-teenagers-prefer/
---
class: center, middle
# Let's open 03-annotating.Rmd

---
class: center, middle
# Part 4: Highlighting Data

---
# Preattentive attributes

.pull-left[
Color and contrast are known as `preattentive attributes`. Our unconscious brain is aware of these kinds of attributes even before we consciously process the content of a graph.

How many 3s are there in this image?
]

.pull-right[
<img src="image/threes-grey.png" width="743" />
]

.footnote[[Storytelling with Data](http://storytellingwithdata.com/)]
---
# How about now?

.pull-left[
You can use color and contrast to highlight aspects of the data. How many 3s are there in this image now? Notice how much quicker you can count them.

That's the power of preattentive attributes!
]

.pull-right[
<img src="image/threes.png" width="743" />
]

.footnote[[Storytelling with Data](http://storytellingwithdata.com/)]
---
# Best Practices for Using Color (Stephen Few)

- Use color only when needed to serve a particular communication goal
- Use different colors only when they correspond to differences of meaning in the data.
- Use soft, natural colors to display most information and bright and/or dark colors to highlight information that requires greater attention.

.footnote[[Practical Rules for Using Color](http://www.perceptualedge.com/articles/visual_business_intelligence/rules_for_using_color.pdf)]

---
class: center, middle
# Let's open up 04-highlighting.Rmd

---
# Conclusions

Congrats! You're well on your way to learning how to make your figures more accessible.

---
class: center, middle
# Putting it all Together

---
<img src="image/David_H.png" width="1323" />
.footnote[https://t.co/KSGQzaH0Mh]
---
class: center, middle

# Going Farther
---
# `ggplot2` flipbook

Good examples for styling your plots!

https://evamaerey.github.io/ggplot_flipbook/ggplot_flipbook_xaringan.html

- [Arctic Ice](https://evamaerey.github.io/ggplot_flipbook/ggplot_flipbook_xaringan.html#226)
- [Flipping Seats](https://evamaerey.github.io/ggplot_flipbook/ggplot_flipbook_xaringan.html#302)
- [Milk Cows](https://evamaerey.github.io/ggplot_flipbook/ggplot_flipbook_xaringan.html#354)
---
# CSE631 Principles and Practice of Data Visualization

- [Excellent course](https://cslu.ohsu.edu/~bedricks/courses/cs631/) that will have you learning many more principles of visualizations
- More work with R!
---
class: center, middle
# Please fill out our survey!
http://bit.ly/st_survey

---
# References

- [Storytelling with Data](http://www.storytellingwithdata.com/books)
- [Happy Days Jumping the Shark (Tableau Link)](https://t.co/KSGQzaH0Mh)
- [`ggplot2` flipbook](https://evamaerey.github.io/ggplot_flipbook/ggplot_flipbook_xaringan.html)
- [Alison Hill: Take a Sad Plot and Make it Better](https://alison.rbind.io/talk/2018-ohsu-sad-plot-better/)
- Slides are done with xaringan/xaringanthemer

---
# Keep in Touch

laderast at ohsu.edu 
 https://laderast.github.io 
 [tladeras](https://twitter.com/tladeras)