Data visualization

# Truth, beauty, and data

**andhs.co/church-dataviz**

.light.small[Andrew Young School of Policy Studies 
Georgia State University 
July 7, 2020]

]

---

# Plan for today

.box-2.medium[Truth and beauty]

.box-3.medium[Data and beauty]

.box-5.medium[Design and beauty]

.box-7.medium[What next?]

---

layout: false
name: facts-truth-beauty
class: center middle section-title section-title-2 animated fadeIn

# Truth and beauty

---

---

# What is truth?

.box-inv-2[“The first and fundamental principle of our holy religion is, that we believe that we have a right to embrace all, and every item of truth”]

.pull-right.center[
<figure>
 <img src="img/js.jpg" alt="Joseph Smith" title="Joseph Smith" width="70%">
</figure>
]

???

https://upload.wikimedia.org/wikipedia/commons/8/86/Joseph_Smith%2C_Jr._portrait_owned_by_Joseph_Smith_III.jpg

---

# What is truth?

.box-inv-2.medium[Core principles of the universe?]

.box-inv-2.medium[Underlying trends in society?]

.box-inv-2.medium[Something transcendental?]

.box-inv-2.medium[Reality?]

---

# How do we find truth?

.box-inv-2.medium[Science!]

.center[
<figure>
 <img src="img/ndgt-scientism.png" alt="Neil deGrasse Tyson on science" title="Neil deGrasse Tyson on science" width="70%">
</figure>
]

???

[Original tweet](https://twitter.com/neiltyson/status/345551599382446081)

---

# But wait!

.box-inv-2.medium.sp-after[Beware of scientism!]

.box-2["… promotion of science as the best or only objective means by which society should determine normative and epistemological values"]

???

[Wikipedia](https://en.wikipedia.org/wiki/Scientism)

---

# Science is not the only way

.box-inv-2.medium[Art]

.box-inv-2.medium[Music]

.box-inv-2.medium[Literature]

.box-inv-2.medium[Religion]

.box-inv-2.medium[Nature]

???

None of these methods deal with raw facts, though!

---

# Nothing here is factual…

.pull-left[
<figure>
 <img src="img/les-mis.jpg" alt="ZZZ" title="ZZZ" width="100%">
 <figcaption>Cosette</figcaption>
</figure>
]

.pull-right[
<figure>
 <img src="img/King_Lear.jpg" alt="King Lear" title="King Lear" width="100%">
 <figcaption>King Lear</figcaption>
</figure>
]

]

.pull-right[
<figure>
 <img src="img/beethoven.png" alt="ZZZ" title="ZZZ" width="100%">
 <figcaption>Beethoven's 9th symphony</figcaption>
</figure>
]

???

Sources:

- [Cosette](https://commons.wikimedia.org/wiki/File:Cosette-sweeping-les-miserables-albert-bellenger-1886.jpg)
- [King Lear](https://commons.wikimedia.org/wiki/File:King_Lear.jpg)
- [Beethoven](https://commons.wikimedia.org/wiki/File:Beethoven,_Symphony_No._9,_finale,_bars_116-123.png)

---

.box-2.huge[Facts ≠ truth]

---

.box-2.large.sp-after[Where does truth come when there are no facts?]

.box-inv-2.huge[Beauty]

---

---

# Beauty in science

.pull-left.center[

]

.box-inv-2.medium[This is also true for science and math and other more factual realms]

---

# Rhetoric and beauty

.box-inv-2.tiny.sp-after[Logos&emsp;•&emsp;Lexis]

.box-inv-2.sp-after[Content&emsp;•&emsp;Form]

.box-2.medium[Truth&emsp;•&emsp;Beauty]

???

One is the underlying metaphysical meaning—what the author or artist intends. The other is *how* they convey that message—what it looks like, what they used, etc.

---

# Content + form

.box-inv-2.large[Art is how we translate core, essential **content** (or truth!) to different **forms** for specific **audiences**.]

---

# Truth is beautiful

.box-inv-2.medium[Truth ≠ facts]

.box-inv-2.medium.sp-after[Truth comes from aesthetic combination of content and form]

.box-2.medium[Facts require beauty to be true]

---

layout: false
name: data-truth-beauty
class: center middle section-title section-title-3 animated fadeIn

# Data and beauty

---

---

# Just show me the data!

.pull-left-3.small-code[

```r
head(my_data, 10)
```

```
## # A tibble: 10 x 2
## x y
## <dbl> <dbl>
## 1 55.4 97.2
## 2 51.5 96.0
## 3 46.2 94.5
## 4 42.8 91.4
## 5 40.8 88.3
## 6 38.7 84.9
## 7 35.6 79.9
## 8 33.1 77.6
## 9 29.0 74.5
## 10 26.2 71.4
```
]

.pull-middle-3.small-code[

```r
mean(my_data$x)
```

```
*## [1] 54.26327
```

```r
mean(my_data$y)
```

```
*## [1] 47.83225
```

```r
cor(my_data$x, my_data$y)
```

```
*## [1] -0.06447185
```
]

&nbsp;

&nbsp;

---

# oh no

.center[
<figure>
 <img src="img/DinoSequentialSmaller.gif" alt="Datasaurus Dozen" title="Datasaurus Dozen" width="100%">
 <figcaption><a href="https://www.autodeskresearch.com/publications/samestats" target="_blank">The Datasaurus Dozen</a></figcaption>
</figure>
]

---

# Raw data is not enough

.box-inv-3.small[Each of these has the same mean, standard deviation, variance, and correlation]

---

# Humans love patterns

.center[
<figure>
 <img src="img/pattern-processing.png" alt="Pattern processing" title="Pattern processing" width="100%">
 <figcaption><a href="https://doi.org/10.3389/fnins.2014.00265" target="_blank">https://doi.org/10.3389/fnins.2014.00265</a></figcaption>
</figure>
]

---

# (Sometimes we love them too much)

---

# Beauty is necessary to see patterns

]

.pull-right.center[

]

---

# What makes a great visualization?

.box-inv-3["Graphical excellence is the **well-designed presentation of interesting data**—a matter of substance, of statistics, and of design … [It] consists of complex ideas communicated with clarity, precision, and efficiency. … [It] is that which **gives to the viewer the greatest number of ideas in the shortest time with the least ink in the smallest space** … [It] is nearly always multivariate … And graphical excellence requires **telling the truth about the data**."]

.box-3.tiny[Edward Tufte, *The Visual Display of Quantitative Information*, p. 51]

---

# What makes a great visualization?

.box-inv-3.medium[Good aesthetics]

.box-inv-3.medium[No substantive issues]

.box-inv-3.medium[No perceptual issues]

.box-inv-3.medium[Honesty + good judgment]

.box-3.tiny[Kieran Healy, *Data Visualization: A Practical Introduction*]

---

# What's wrong?

.center[
<figure>
 <img src="img/pie-genus.png" alt="Distribution of genus" title="Distribution of genus" width="60%">
</figure>
]

???

- Aesthetic issues
- Substantive issues
- Perceptual issues
- Honesty + judgment issues

---

# What's wrong?

.center[
<figure>
 <img src="img/changing-face-of-america.png" alt="Changing face of America" title="Changing face of America" width="45%">
</figure>
]

???

- Aesthetic issues
- Substantive issues
- Perceptual issues
- Honesty + judgment issues

---

# What's wrong?

.center[
<figure>
 <img src="img/how-stuff-works-recidivism.png" alt="Recidivism pie chart" title="Recidivism pie chart" width="45%">
</figure>
]

???

- Aesthetic issues
- Substantive issues
- Perceptual issues
- Honesty + judgment issues

---

# What's right?

]

???

- Aesthetic issues
- Substantive issues
- Perceptual issues
- Honesty + judgment issues

.pull-right.center[

<figure>
 <img src="img/bergstrom-tweet.png" alt="Carl Bergstrom tweet" title="Carl Bergstrom tweet" width="80%">
 <figcaption><a href="https://twitter.com/CT_Bergstrom/status/1235865328074153986" target="_blank">Thread by Carl T. Bergstrom</a></figcaption>
</figure>

]

???

[Carl Bergstrom on the flatten the curve visualization](https://twitter.com/CT_Bergstrom/status/1235865328074153986)

---

layout: false
class: bg-full bg-y-75
background-image: url("img/napoleon-retreat.jpg")

???

Source: [Wikipedia](https://en.wikipedia.org/wiki/File:National_Museum_in_Poznan_-_Przej%C5%9Bcie_przez_Berezyn%C4%99.JPG)

---

---

# Long distance!

.center[
<figure>
 <img src="img/napoleon-google-maps.png" alt="Moscow to Vilnius" title="Moscow to Vilnius" width="80%">
 <figcaption>Moscow to Vilnius</figcaption>
</figure>
]

---

# Very cold!

---

# Lots of people died!

---

???

Source: [Wikimedia Commons](https://upload.wikimedia.org/wikipedia/commons/2/29/Minard.png)

---

name: truth-beauty-design
class: center middle section-title section-title-5 animated fadeIn

# Design and beauty

---

---

# Why even care about design?

.center[
<figure>
 <img src="img/why-design.png" alt="Why care about design?" title="Why care about design?" width="100%">
</figure>
]

---

# Beauty, stories, and truth

.box-inv-5.sp-after[**Beauty** shapes that form.]

.box-5.medium[There are aesthetic principles we can follow to create beauty—and truth.]

---

# Principles of design

.box-inv-5.medium[There are thousands of books and centuries of debate and theory about what makes good design.]

.box-5.medium[The easiest and most memorable distillation of these principles is **CRAP**.]

---

# CRAP

]

.pull-right.center[
<figure>
 <img src="img/williams-design-book.jpg" alt="Robin Williams's The Non-Designers Design Book" title="Robin Williams's The Non-Designers Design Book" width="65%">
 <figcaption>Buy this book and your life will change forever.</figcaption>
</figure>
]

---

---

# Contrast

.box-inv-1.large[If two items are not exactly the same, make them different. Really different.]

.box-1.large[**Don't be a wimp.**]

---

# Typographic contrast: Family

.box-1.small[Serif + Sans serif&emsp;&emsp;Script + Serif&emsp;&emsp;Slab + Sans serif&emsp;&emsp;etc.]

<table>
 <tr>
 <td class="cell-left">Serif</td>
 <td class="cell-left" style="font-family: 'EB Garamond'">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Sans serif</td>
 <td class="cell-left" style="font-family: 'Oswald'">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Slab serif</td>
 <td class="cell-left" style="font-family: 'Alfa Slab One'">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Script</td>
 <td class="cell-left" style="font-family: 'Dancing Script'">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Monospaced&emsp;</td>
 <td class="cell-left" style="font-family: 'Inconsolata'">Sphinx of black quartz, judge my vow</td>
 </tr>
</table>

---

# Typographic constrast: Family

.pull-left.frame[

Here’s a heading

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.

]

.pull-right.frame[

Here’s a heading

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.

]

---

# Typographic contrast: Weight

.box-1.small[Bold + Regular&emsp;&emsp;Regular + Extra light&emsp;&emsp;Black + Light&emsp;&emsp;etc.]

<table>
 <tr>
 <td class="cell-left">Extra light&emsp;</td>
 <td class="cell-left" style="font-family: 'Source Sans Pro'; font-weight: 200;">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Light</td>
 <td class="cell-left" style="font-family: 'Source Sans Pro'; font-weight: 300;">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Regular</td>
 <td class="cell-left" style="font-family: 'Source Sans Pro'; font-weight: 400;">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Semi bold</td>
 <td class="cell-left" style="font-family: 'Source Sans Pro'; font-weight: 600;">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Bold&emsp;</td>
 <td class="cell-left" style="font-family: 'Source Sans Pro'; font-weight: 700;">Sphinx of black quartz, judge my vow</td>
 </tr>
 <tr>
 <td class="cell-left">Black</td>
 <td class="cell-left" style="font-family: 'Source Sans Pro'; font-weight: 900;">Sphinx of black quartz, judge my vow</td>
 </tr>
</table>

---

# Typographic constrast: Weight

.pull-left.frame[

Here’s a heading

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.

]

.pull-right.frame[

Here’s a heading

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.

]

---

# Size contrast

&nbsp;

---

# Color contrast

.box-1.tiny[[color.adobe.com](https://color.adobe.com)]

.pull-left[
<figure>
 <img src="img/triad.png" alt="Triad" title="Triad" width="100%">
 <figcaption>Triad</figcaption>
</figure>

<figure>
 <img src="img/monochromatic.png" alt="Monochromatic" title="Monochromatic" width="100%">
 <figcaption>Monochromatic</figcaption>
</figure>
]

.pull-right[
<figure>
 <img src="img/complementary.png" alt="Complementary" title="Complementary" width="100%">
 <figcaption>Complementary</figcaption>
</figure>

<figure>
 <img src="img/complementary-split.png" alt="Split complementary" title="Split complementary" width="100%">
 <figcaption>Split complementary</figcaption>
</figure>
]

---

# Color contrast

.center[
<figure>
 <img src="img/palette-atl-extracted.png" alt="Extracted colors" title="Extracted colors" width="50%">
 <figcaption>Colors extracted from an image at <a href="https://color.adobe.com">color.adobe.com</a></figcaption>
</figure>
]

???

Photo by [Brad Huchteman](https://unsplash.com/photos/sR1Kz2auNJE) on Unsplash

---

# Usability and accessibility

.box-inv-1.medium[Perceptually uniform colors]

.box-1[Values close to each other use similar colors, and values far from each other use different colors]

.box-inv-1.medium[Colorblind-safe colors]

---

# Perceptually uniform colors

.pull-left[
<figure>
 <img src="img/palettes-typical.png" alt="Typical palettes" title="Typical palettes" width="100%">
 <figcaption>Traditional palettes vs. viridis</figcaption>
</figure>
]

.pull-right[
<figure>
 <img src="img/palettes-deuteranopic.png" alt="Deuteranopic palettes" title="Deuteranopic palettes" width="100%">
 <figcaption>Traditional palettes vs. viridis as seen with deuteranopia</figcaption>
</figure>
]

---

# Perceptually uniform colors

.pull-left-3[
<figure>
<img src="index_files/figure-html/ga1-1.png" width="360" style="display: block; margin: auto;" />

<figcaption>Georgia counties filled by area, rainbow palette (NOT GOOD)</figcaption>
</figure>
]

.pull-middle-3[
<figure>
<img src="index_files/figure-html/ga2-1.png" width="360" style="display: block; margin: auto;" />

<figcaption>Georgia counties filled by area, viridis::viridis palette</figcaption>
</figure>
]

.pull-right-3[
<figure>
<img src="index_files/figure-html/ga3-1.png" width="360" style="display: block; margin: auto;" />

<figcaption>Georgia counties filled by area, viridis::inferno palette</figcaption>
</figure>
]

---

---

# Repetition

&nbsp;

.box-4.large[Repeat some aspect of the design throughout the entire piece.]

---

# Things to repeat

.box-4.medium[Colors]

.box-4.medium[Fonts]

.box-inv-4.sp-after[Families, weights, sizes]

.box-4.medium[Graphical elements]

.box-4.medium[Alignments]

---

layout: false
class: bg-full
background-image: url("img/oxfam-annual-report.jpg")

???

Source: Pages 5 and 9 from Oxfam's 2019 Annual Report/: https://assets.oxfamamerica.org/media/documents/2019_Oxfam_America_Annual_Report_web.pdf

If you want to see generally great examples of public sector designs, search for "<big charity or think tank name> annual report", like "red cross annual report" or "urban institute annual report"

---

---

# Alignment

&nbsp;

.box-1.large[Every item should have a visual connection with something else on the page.]

---

# Alignment

.pull-left[
<figure>
 <img src="img/alignment-bad.png" alt="Bad alignment" title="Bad alignment" width="100%">
 <figcaption>Bad alignment</figcaption>
</figure>
]

.pull-right[
<figure>
 <img src="img/alignment-good.png" alt="Good alignment" title="Good alignment" width="100%">
 <figcaption>Good alignment—everything is connected to something</figcaption>
</figure>
]

???

Source: Robin Williams, *The Non-Designer's Design Book*

---

# Alignment + repetition

.pull-left[
<figure>
 <img src="img/alignment-bad-annotated.png" alt="Bad alignment" title="Bad alignment" width="100%">
 <figcaption>4 vertical alignments; 3 horizontal alignments</figcaption>
</figure>
]

.pull-right[
<figure>
 <img src="img/alignment-good-annotated.png" alt="Good alignment" title="Good alignment" width="100%">
 <figcaption>1 shared vertical alignment; 2 horizontal alignments</figcaption>
</figure>
]

---

# Alignment + contrast

.pull-left.frame[

Chapter 1

Here’s a title

Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua.

]

.pull-right.frame[

Chapter 1

Here’s a title

Lorem ipsum dolor sit amet, con&shy;sec&shy;tetur adi&shy;pis&shy;icing elit, sed do eiusmod tempor in&shy;ci&shy;di&shy;dunt ut labore et dolore magna aliqua.

]

---

---

# Proximity

&nbsp;

.box-inv-4.large[Group related items together.]

---

# Proximity

.box-inv-4[Use white space, color, location, contrast, repetition, alignment, etc. to make visually distinct groupings]

.pull-left[
<figure>
 <img src="img/proximity-bad.png" alt="Bad proximity" title="Bad proximity" width="100%">
 <figcaption>Bad proximity; no logical groupings</figcaption>
</figure>
]

.pull-right[
<figure>
 <img src="img/proximity-good.png" alt="Good proximity" title="Good proximity" width="100%">
 <figcaption>Good proximity; information visually grouped</figcaption>
</figure>
]

???

Source: Robin Williams, *The Non-Designer's Design Book*

---

---

# CRAP review

<figure>
 <img src="img/review-contrast.png" alt="Contrast" title="Contrast" width="100%">
</figure>
]

<figure>
 <img src="img/review-repetition.png" alt="Repetition" title="Repetition" width="100%">
</figure>
]
]

<figure>
 <img src="img/review-alignment.png" alt="Alignment" title="Alignment" width="100%">
</figure>
]

<figure>
 <img src="img/review-proximity.png" alt="Proximity" title="Proximity" width="100%">
</figure>
]
]

---

layout: false
name: moral-story
class: center middle section-title section-title-7 animated fadeIn

# What next?

---

---

# Skills to develop

.box-inv-7.medium.sp-after[Visualization]

.box-inv-7.medium.sp-after[Design]

.box-inv-7.medium[Storytelling]

---

# Visualization software

.center[
<figure>
 <img src="img/complexity-chart@4x.png" alt="Complexity chart" title="Complexity chart" width="65%">
</figure>
]

---

# Visualization resources

.small[
- Andrew Heiss, ["Data Visualization with R"](https://datavizm20.classes.andrewheiss.com/) (free online course)
- Kieran Healy,[ *Data Visualization for Social Science: A practical introduction with R and ggplot2*](http://socviz.co/)
- Claus Wilke, [*Fundamentals of Data Visualization*](https://serialmentor.com/dataviz/)
- Alberto Cairo, *The Truthful Art: Data, Charts, and Maps for Communication* 
- Stephanie D. H. Evergreen, *Effective Data Visualization: The Right Chart for the Right Data*
- Dona M. Wong, *The Wall Street Journal Guide to Information Graphics: The Dos and Don’ts of Presenting Data, Facts, and Figures* 
- Hadley Wickham and Garrett Grolemund, *R for Data Science: Import, Tidy, Transform, Visualize, and Model Data*
- Alberto Cairo, *The Functional Art: An Introduction to Information Graphics and Visualization*
]

---

# Design resources

.center[
<figure>
 <img src="img/williams-design-book.jpg" alt="Robin Williams's The Non-Designers Design Book" title="Robin Williams's The Non-Designers Design Book" width="30%">
 <figcaption>Buy this book and your life will change forever.</figcaption>
</figure>
]

---

# Design resources

## Accessibility

- [**Vischeck**](http://www.vischeck.com/vischeck/vischeckImage.php): Simulate how your images look for people with different forms of colorblindness (web-based)
- [**Color Oracle**](http://colororacle.org/index.html): Simulate how your images look for people with different forms of colorblindness (desktop-based, more types of colorblindness)

## Fonts

- [**Google Fonts**](https://fonts.google.com/): Huge collection of free, well-made fonts.
- [**The Ultimate Collection of Google Font Pairings**](https://www.reliablepsd.com/ultimate-google-font-pairings/): A list of great, well-designed font pairings from all those fonts hosted by Google (for when you're looking for good contrasting or complementary fonts).
]

## Colors

- [**Adobe Color**](https://color.adobe.com): Create, share, and explore rule-based and custom color palettes.
- [**ColourLovers**](http://www.colourlovers.com/): Like Facebook for color palettes.
- [**viridis**](https://cran.r-project.org/web/packages/viridis/vignettes/intro-to-viridis.html): Percetually uniform color scales.
- [**ColorBrewer**](http://colorbrewer2.org/): Sequential, diverging, and qualitative color palettes that take accessibility into account.
- [**Colorgorical**](http://vrl.cs.brown.edu/color): Create color palettes based on fancy mathematical rules for perceptual distance.
- [**Colorpicker for data**](http://tristen.ca/hcl-picker/): More fancy mathematical rules for color palettes ([explanation](https://www.vis4.net/blog/posts/avoid-equidistant-hsv-colors/)).
- [**iWantHue**](http://tools.medialab.sciences-po.fr/iwanthue/): Yet another perceptual distance-based color palette builder.
- [**Photochrome**](https://photochrome.io/): Word-based color palettes.
- [**PolicyViz Design Color Tools**](https://policyviz.com/better-presentations/design-resources/design-color-tools/): Large collection of useful color resources

]
]

---

# Storytelling

.pull-left.small[

## Videos

- Ben Wellington, ["Making data mean more through storytelling"](https://www.youtube.com/watch?v=6xsvGYIxJok)
- Jonathan Schwabish, ["Better data communication"](https://vimeo.com/230757062)

]

.pull-right.small[

## Books

- Cole Nussbaumer Knaflic, *Storytelling with Data: A Data Visualization Guide for Business Professionals*
- Alan Alda, *If I Understood You, Would I Have This Look on My Face? My Adventures in the Art and Science of Relating and Communicating*
- Nancy Duarte, *Resonate: Present Visual Stories That Transform Audiences*
]

---

# Moral of the story

.box-inv-7.medium.sp-after[Facts require beauty to be true]

.box-inv-7.medium.sp-after[Design + stories = truth]

.box-inv-7.medium[You can (and should) develop these skills!]