GEI1002

Computers and
the humanities

Introductory Lecture
Concepts and module structure
1
GEI1002
GET1030

Data is everywhere

Data is used everywhere. It

is important for everyone to
be a thoughtful producer
and consumer of data.

2
GEI1002
GET1030

Data literacy

• Ability to obtain and analyze data

(technical skills)
• Asking questions about the ways data
was obtained and analyzed (interpretive
skills)

3
GEI1002
GET1030

This module

• We will focus on the analysis and

creation of data visualizations.
• The lectures and tutorials will use
examples from the arts and culture
(film, literature, etc.) but you can use
any topic from the humanities or social
sciences for your projects.

4
GEI1002

Objectives of the module

- To get you thinking about interdisciplinary work

- To provide you with some general skills that
are useful for many other fields
- To give you a foundation from which you can
learn on your own

5
GEI1002

General skills you will learn

- How to evaluate datasets and

visualizations
- How to create a wide range of
visualizations

6
GEI1002

Two interconnected components

Doing stuff: learning a bit of survival coding to

understand the computational process of dataviz.

Thinking: stepping back and reflecting on what we have

done.

These two steps are not distinct, they feed each other.

This double focus on doing and thinking shapes our

assignments and the structure of the module.

7
GET1030

Teaching modes

We follow a flipped classroom model, all lectures are

delivered through video. In the tutorials we will do two
things:

● Discussions in small groups: based on readings

● Coding labs: learn how to create exploratory data
visualizations in Python (no coding experience
required)

8
 GEI1002
Computers and the humanities

Introductory Lecture
Part II. What is data?
9
GEI1002
GET1030

What is data?

- Systematic observations about a

phenomenon
- Variables and values
- Example: a dataset of books
- What are some possible variables?
- Author

- Genre

- Number of pages

- Average rating

10
GEI1002
GET1030

Types of data

- Quantitative and categorical data

- This refers to the possible values a variable
can take
- Quantitative data
- Number of pages: 292, 165
- Rating: 3.6, 4.1
- Categorical data
- Genre: Horror, Sci-Fi, Mystery, Romance
- Author: Arthur Conan Doyle, Agatha Christie
11
GEI1002
GET1030

Subjectivity in data

- Subjective vs objective data

- Ratings for The Adventures of Sherlock
Holmes, from goodreads.com

- Is this data objective?

- Remember that quantitative data is not
necessarily objective
12
GEI1002
GET1030

Subjectivity in data

- Subjective data is not useless, but you need to be careful

about what claims you can make with it.
- These two statements are not the same:
- “The Adventures of Sherlock Holmes is a good book”
-“The Adventures of Sherlock Holmes has a high
average rating on goodreads.com”
- Can you use the goodreads.com data to claim that a
book is good? It depends on the context.

13
GEI1002
GET1030

Subjectivity in data

- And “categorical” data is not necessarily very

subjective
- Can you think of an example?

14
GEI1002
GET1030

Authors as “categories”

15
Source: https://towardsdatascience.com/all-the-authors-around-us-an-analytical-look-into-goodreads-authors-
dataset-part-one-61697721e58e
GEI1002
GET1030

Decisions shape the data

- A dataset is always shaped by the decisions

of the people who made it
- Lets imagine a dataset with book genres

16
GEI1002
GET1030

Decisions shape the data

- What about a book that fits into two

categories?
- Shall we list both categories?

- This constraints the types of analysis you can

make
17
GEI1002
GET1030

Definitions

- Different definitions will lead to different

datasets (Thorp, 2021)
- A good data project describes the
definitions used, and provides links to the
sources

18
GEI1002
GET1030

Definitions

19
https://works.periscopic.com/unicef-child-violence/#all&criteria=1
GEI1002
GET1030

Limitations

- There are always limitations in the ways

data is collected
- A good data project acknowledges these
limitations

20
GEI1002
GET1030

Limitations

Covid-19 vaccinations by country

https://www.nytimes.com/interactive/2021/world/covid-vaccinations-tracker.html
21
GEI1002
GET1030

Errors in data

- Even the best intended datasets might contain errors

https://www.bbc.com/news/technology-54423988
22
GEI1002
GET1030

Context

- Pay attention to who made the data and why.

- For example, what is the objective of
goodreads.com?
- How is it different from the objective of
Unicef’s “#ENDviolence”?

23
GEI1002
GET1030

“Tidy” data

- For this module, we will mostly use

spreadsheets with “tidy” data.
- The variables are the columns.
- There is one observation per row.

24
GEI1002
GET1030

“Tidy” data

25
GEI1002
GET1030

Summary

- Data refers to systematic observations

- A dataset includes values, variables and observations
- In “tidy” datasets there is one observation per row and the columns
represent the variables
- Variables can be quantitative or categorical, and they can be more
subjective or more objective
- But no dataset is fully objective
- Pay attention to the context of a data project:
- Objectives
- Definitions
- Limitations
- Assumptions
- A good data project clearly states its limitations and sources.

26
 GEI1002
Computers and the humanities

Introductory Lecture
Part III. Data visualizations
27
GEI1002
GET1030

What is data visualization?

- Representation of data using visual conventions:

shapes, colors, distances, symbols.
- Fundamental activity for making sense of data.
- Data visualization ~ infographics, graphs, charts.

28
GEI1002
GET1030

Quantitative description

- In this module, we will focus on “quantitative

descriptions”, sometimes called “exploratory
data analysis”.
- We will not try to prove hypotheses, as this is
a more advanced topic.

29
GEI1002
GET1030

Quantitative description

Number of art biennales around the world over time

Tifentalle and Manovich (2020) 30

GEI1002
GET1030

Choosing a data visualization

- What task is the visualization aimed to help

people achieve? (Cairo 2012)

31
GEI1002
GET1030

Choosing a data visualization

From Cairo (2012)

What tasks does this visualization enable?

32
GEI1002
GET1030

Choosing a data visualization

Here is the same data, but visualized as a barchart.

From Cairo (2012)

33
GEI1002
GET1030

Choosing a data visualization

- Is the visualization suitable to its context?

- Identifying this is an art and a science.

34
GEI1002
GET1030

Choosing a data visualization

Alberto Cairo (2012)’s guide for choosing a

visualization, based on Cleveland and McGill’s
elementary perceptual tasks (see Files for full 35
PDF).
GEI1002
GET1030

Good data visualizations?

- There are many “rules” and “best practices”

- Often they are not based on empirical
evidence
- It is important to pay attention to context
- What task is the visualization aimed to help
people achieve?
- Your argument is more important than getting
things “right”.

36
GEI1002
GET1030

Summary

-Representation of data using visual

conventions: shapes, colors, distances,
symbols.
- Key question: what task is the visualization
aimed to help people achieve?
- Your argument is more important than getting
things “right”.

37
 GEI1002
Computers and the humanities

Introductory Lecture
Part IV. Case studies: using data to study culture
38
GEI1002
GET1030

Two types of mediation

-World to Data
-Data to Image
See Gray et al (2016) for more on this.

*always pay attention to the decisions made by

the people who created the data or the
visualizations
*if you are the creators, justify your choices!
39
GEI1002
GET1030

Two types of mediation

A research project: How many current authors

are people of color (PoC)?

40
GEI1002

The publishing industry

Source: “Redlining Culture” by Richard Jean So.

https://www.nytimes.com/interactive/2020/12/11/opinion/culture/diversity-publishing-industry.html 41
GEI1002
GET1030

Two types of mediation

First, we gathered a list of English-language fiction books published between

1950 and 2018 […]

We also constrained our search to books released by some of the most prolific
publishing houses […] After all that we were left with a dataset containing
8,004 books, written by 4,010 authors.

To identify those authors’ races and ethnicities, we worked alongside three

research assistants, reading through biographies, interviews and social media
posts. Each author was reviewed independently by two researchers. If the
team couldn’t come to an agreement about an author’s race, or there simply
wasn’t enough information to feel confident, we omitted those authors’ books
from our analysis. By the end, we had identified the race or ethnicity of 3,471
authors.

42
GEI1002

The publishing industry

Source: “Redlining Culture” by Richard Jean So (2020). 43

GEI1002
GET1030

Data and Culture

We will see many different types of data and

visualizations. But we must always ask:
- How was a phenomenon represented as data?
- How was this data visualized?
- What decisions were made?
- Do they make sense within their context?

44
GEI1002

Sucess in art

Network of 12,238 exhibition

venues for artists

45
Fraiberger et al (2018)
GEI1002

Phototrails

Locals and tourists in New York.

The visualization compares
locations of photos uploaded to
Flickr and Picasa. Blue pictures
are by locals. Red pictures are by
tourists. Yellow pictures might be
by either.

46
Hochman and Manovich (2013)
https://firstmonday.org/ojs/index.php/fm/article/view/4711/3698
 GEI1002
Computers and the humanities

Introductory Lecture
Part V. Structure and Assignments
47
GEI1002

Module Structure

Concepts
1 Data, computation and the humanities
2 What is data?
3 Visualizing data

Tools
4 Gentle introduction to data visualization in Python
5 Visualizing text
6 Tools for visualizing text
7 Visualizing networks
8 Tools for visualizing networks
9 Visualizing geographical data
10 Tools for visualizing geographical data

Looking beyond
11 Computation and society All lectures will be video
based
12 Group project consultations
13 Group project presentation slam
48
GEK 2050
GEI1002

Tutorial Sessions

Concepts Tutorial Sessions

1 Data, computation and the humanities
2 What is data? #1 Concepts (Week 3)
3 Visualizing data

Tools
4 Data visualization in Python #2 Python visualizations (Week 5)
5 Visualizing text
6 Tools for visualizing text #3 Text (Week 7)

7 Visualizing networks
8 Tools for visualizing networks #4 Networks (Week 9)

9 Visualizing spatial data #5 Spatial visualizations (Week 11)

10 Tools for visualizing spatial data

Looking beyond
11 Computation and society
12 Project consultations [no lecture] 49
13 Presentation slam
GEK 2050
GEI1002

A note on programming

There’s a very gentle introduction to programming in Week 4.

This is not a full-fledged programming module, and I hope
the simple exercises will get you interested in learning more
about programming (if you haven’t already done this).

We will only learn to load Excel files into python and to

visualize them through simple commands (using interactive
Jupyter notebooks).

50
GEK 2050
GEI1002

Tools for the module

None of these require programming.

Voyant Tools for textual analysis http://voyant-

tools.org/

Google maps for geographical visualizations

http://maps.google.com

Gephi for network analysis

https://gephi.org/
51
GEK 2050
GEI1002

ASSESSMENT

See the PDF for details on the assessment schedule and

description of the assignments.

52
GEK 2050
GEI1002

REFERENCES

Cairo, Alberto. The Functional Art: An Introduction to Information Graphics and Visualization. 1st edition. Berkeley, California: New
Riders, 2012.

Thorp, Jer. Living in Data: A Citizen’s Guide to a Better Information Future. New York: MCD, 2021.

Gray, Jonathan et al. ‘Ways of Seeing Data: Toward a Critical Literacy for Data Visualizations as Research Objects and Research
Devices’. In Innovative Methods in Media and Communication Research, edited by Sebastian Kubitschko and Anne Kaun, 227–
51. Cham: Springer International Publishing, 2016.

So, Richard Jean. Redlining Culture: A Data History of Racial Inequality and Postwar Fiction. New York: Columbia University
Press, 2020.

Fraiberger, Samuel P., Roberta Sinatra, Magnus Resch, Christoph Riedl, and Albert-László Barabási. 2018. “Quantifying
Reputation and Success in Art.” Science 362 (6416): 825–29. https://doi.org/10.1126/science.aau7224.

Hochman, Nadav, and Lev Manovich. 2013. “Zooming into an Instagram City: Reading the Local through Social Media.” First
Monday. https://firstmonday.org/ojs/index.php/fm/article/view/4711/3698

53