Visualizing Complexity: Modular Information Design Handbook 9783035625066, 9783035625042

Citation preview

Visualizing Complexity

Darjan Hil Nicole Lachenmeier

Visualizing Complexity

Modular Information Design Handbook

Birkhäuser Basel

Contents

6 8 11

Preface How the book is structured The MID system

1 Data dimensions

15 17

1A From text to data 1B From data to data cube

31

41

2 Diagrammatical dimensions

43 59 67

2A Quantity 2B Position 2C Relationship

77

3 Visual dimensions

79 87 97 107 115 123

3A Color 3B Shape 3C Line 3D Pattern 3E Contour 3F Isotype

133 135 147

154

4 Structuring dimensions 4A Structuring by sorting 4B Structuring by grouping

5 The variety of ­multidimensional ­visualizations

156

Overview of all 80 elements

158

How the MID system works

160

Mapping four data dimensions — instructions

162

How sorting works

164

26 multidimensional ­visualizations

217 218 220

6 Appendix

The authors Further reading, sources, and tools

Preface

The title of this book, “Visualizing Complexity”, contains the word complexity, a frequently used term, which quite fittingly describes current thinking. We live in a time when, thanks to computers and an infinite amount of information, people are doing more and more tasks at the same time, taking on different roles and switching very quickly between subjects. In this context, we understand complexity to mean that an issue is impacted by many factors, all of them interwoven, and is obscure and difficult to grasp. But why visualize complexity? Quite simply because we firmly believe that many a seemingly complex issue can be better processed when it is externalized visually and can be viewed from some distance. Initially, it is not about the quality of the visualization, but simply about the process of getting something out of one’s head onto the page (digital or analog) and thereby making it communicable. In the context of this book, we understand issues to be data as a general concept. Whether we are dealing with text or numbers, in our opinion both are raw material, i.e. a data set to be processed. Both can be considered in terms of quantity or quality, and can be processed visually. In this context it is important to be aware that any consideration of an issue assumes a certain perspective and tells an associated story. For each set of data there are various approaches and perspectives. Therein lies the complexity mentioned above. The aim is not to simplify the complexity, but to enable the reader to skilfully navigate through the different perspectives of the topic. To facilitate this navigation, we have developed a method called modular information design; MID for short. In this system, all elements, the smallest units of the system, are made combinable in modules similar to the periodic table system employed in chem­ istry. It is important to note that in order to facilitate simplified orientation in the MID system, elements are grouped into higherlevel dimensions. Dimensions are units used to organize elements at a certain level.

6

This book is the result of ten years of practical experience in which we enjoyed experimenting and carried out analysis work with the aim of making the little known — to date — but demanding interface discipline of information design more accessible through straight­ forward principles. We are convinced that visualization as a general problem-solving approach helps to create greater transparency and enables wider participation and better decision-making. In addi­ tion, people tend to prefer seeing and reading attractively presented text and figures — and understand them better — than the usual pre­ sentations with their standard diagrams. We want to demonstrate that the world of information design has many beautiful facets. We therefore dare — in an unconventional way — to make connections between disciplines and issues where before there were no obvious ones, the objective being to increase interdisciplinary understanding. This book is intended to be a source of inspiration to all those who work with figures so that they may see data from a new per­ spective and are able to explore design options for the communica­ tion of information. For designers, this handbook can make figures more accessible, and thereby open up new design opportunities for the presentation of information. Generally speaking, we want to convey the joy of working correctly, but also creatively, with text and figures, and we want to motivate people to start their own visual­ ization of complexity. For this it is not necessary to have special talent, or a computer, or special training — but instead a pen and paper, an interest, and a willingness to experiment. We hope that you will enjoy working with this handbook and wish you every success. Darjan Hil and Nicole Lachenmeier

7

How the book is ­structured

This book is structured visually and subdivided into color-coded sections. The first four sections convey the necessary basic understanding and form the basis for the examples of multidimensional visualizations found in the fifth section. It is important to know that the book does not have to be read from front to back, but can be skimmed for inspiration. The structure of the ex­ amples can easily be decoded at the end of the book, and details can be checked in the earlier sections.

Yellow section

In the yellow section, the structure of a double page consists of a data set, an explanation, a title, and an abstract icon as the representa­ tion of the principle explained therein. The yellow section, as well as the entire book, only uses a single small data set that describes a fictitious family history.

Blue, red, green sections

In the blue, red, and green sections, the ele­ ments of the modular information design are ex­ plained on a single page each. An abstract icon represents each principle, which is supple­ mented by an element title and an element ID. In this way, the elements can be traced through­ out the entire book. A simple example in the bottom left-hand corner and a short explanation gives each element a context for practical ­application.

Example section

In this section, each example of data select­ ed from the data set is combined on one page with the elements presented, thus creating a multi­dimensional visualization example. Each page consists of a listing of the data dimensions used and the used elements as icons with the element ID. Step by step, the structure becomes clear with the help of an accurate legend and a descriptive title, as well as a short highlighted text with the specific findings for telling the story from the diagram.

8

Explanation

Data set

Title

Icon

Simple example

Element–ID

Descriptive title

Legend Highlight

9

The MID system

The construction of the visualizations in this book is based on a system. We understand the term system to mean that a number of different components, together with certain principles, form a whole. When applying the principle of deliberately selecting ele­ ments from the modular system and combining them with different dimensions to create a new visualization, we speak of modular information design. The MID system consists of four higher-level modules: the data dimensions (yellow), the diagrammatical dimensions (blue), the visual dimensions (red), and the structuring dimensions (green). Together, the last three dimensions form 80 distinctly identifiable elements. The idea behind the MID system is that by combining a data set with selected elements of the three groups, a large number of multidimensional visualizations can be designed. Using a simple framework, it is possible to encode not only all known diagram types but also a nearly infinite number of new visualizations. This system can be understood as a visual algorithm. But likewise, it is also possible to use MID to analyze and decode existing visualiza­ tions. Multi­dimensional information design does not depend on specific tools. It is possible to use pen and paper and a wide range of different software tools. The fifth chapter contains step-by-step instructions.

11

The MID system

Data dimensions

Structuring dimensions

The data-based dimensions describe the steps associated with the analysis of a data set. Here, the primary concern is not procedures relating to statistics but the essential conceptual considerations as to which questions can be clarified with this data set, which stories can be told, and which perspectives can be ­adopted. From this, the selection of the individual data dimensions is derived, i.e. the columns of a data set.

The structuring dimensions have the greatest influence on the story of a visualization. The question here is: What are the data dimensions used for sorting and grouping a diagram? Just by opting for different ways of sorting it is possible to tell different stories. The sorting also determines the arrangement of the indi­ vidual data elements and therefore has a great influence on the overall shape. On the other hand, grouping brings harmony and order into the format.

Diagrammatical dimensions

Multidimensional visualizations

The diagrammatical dimensions contain all ­elements collected from the various areas of mathematics and information technology.  The common denominator of this collection is that a data set can be translated into a simple diagram in a mathematically correct manner. The diagrams thus created describe either quantities, positions in a system of coordinates, or ­relationships.

Through the multidimensional visualizations, all dimensions are applied together. New dia­ gram variations are created from three or more data dimensions that are mapped onto the elements of the other three dimensions. For the multidimensional visualizations, it is always necessary to have at least four elements.

Visual dimensions

The visual dimensions can be looked upon as the creative dimensions. Here, a great many design principles can be used to create a dia­gram with several data dimensions. Gener­ ally speaking, the visual elements presented here are part of the basic vocabulary of design­ ers. We have subdivided these into the fol­ lowing groups: color, shape, line, pattern, con­ tour, and the special group of isotypes, according to Neurath.1

12

Data dimensions

Diagrammatical dimensions

Visual dimensions

Structuring dimensions

Multidimensional visualizations

13

1 Data dimensions What do we need to know about a data set in order to be able to decide which data dimensions should be used for the visualization? Do we already at this stage need analyses or statistics methods to answer this question? Is it possible to look neutrally at a data set whilst observing a certain distance? Are there intermediate stagesat which visual patterns can be detected in the data set or in the different dimensions of the data set? In this chapter we illustrate approaches that answer these and other similar questions and demonstrate a flowing transition from analysis to visualization. Regarding the conceptual consideration of a data set from its possible access points, a fundamentally open attitude is required. This makes it possible to engage in interdisciplinary cooperation and communication on equal terms.

15

1A From text to data In this chapter we present a data set as a basis for the entire book. It is a text about the Wagner and Brown families, which can be used to extract and visualize data in a step-by-step process. The names of the family members are based on the statistically most common family and first names of the 19 th century in Germanand English-speaking countries. The objective is to illustrate the individual steps required for visualization and to make the possible variations visible, without having to change too many para­meters at once. Please note that the story is fictitious and does not relate to real events. This simple family story makes it possible to understand data easily and establish a connection between them and you.

17

Unstructured running text as a starting point

Running text is intended to give the reader time to bring to mind the respective content in one’s imagination. Written stories feature descriptive details and inspire the readers' power of imagination, encouraging them to mentally dive into their own world. The information in a section is unstructured, without however having a negative effect on how easy it is to understand. Narrators tell their stories in the way their thoughts flow. However, for the purpose of searching or obtaining a quick overview, narrations are less suitable. For example, when searching for Otto Wagner’s date of birth in the text about the entrepreneurial family, one has to scan the text line by line until this is found. Clearly in this case we cannot talk about a quick overview or rapid ­navigation.

18

The story of the entrepreneurial families

The success of the Browag AG company goes back to Hermann Wagner and James Brown. Hermann Wagner was born in Vienna in 1871 (died 1961). During his student years he got to know Vienna-born Emma Moser (1875), who was four years younger (died 1960). In 1900, Hermann and Emma married in Vienna, where the children Otto (1901, died 1924) and Paul (1914, died 2011) were also born. With the help of the Moser family’s financial resources, Hermann was able to set up Wagner Farben GmbH and establish it successfully in Vienna. After finishing school, Otto decided to go to Munich to study and Paul decided to follow in his father Hermann’s footsteps. Otto Wagner died tragically of ­tuberculosis in Munich at the age of 23. In the course of his business, Paul traveled to other countries and, in 1934, whilst in Paris, he met his future wife, ­Elisabeth Brown of Brown Chemicals Inc. The company Brown Chemicals Inc. had been founded in London by James Brown (born 1882, died 1947). He benefited greatly from his marriage to Marie Durand (born 1879, died 1951), who came from a prosperous Parisian family. James built up his business in London as well as in Paris, where the couple spent most of their time. Marie and James had two daughters, Anna (born 1913, died 1996) and Elisabeth (born 1915, died 2014). Anna was born while the family spent some time in London; she grew up there, went to a London boarding school, and spent the rest of her life in London. Her sister Elisabeth grew up in Paris where she lived with her parents; from an early age, she joined papa James’s company, where she later met Paul Wagner. The wedding between Paul and Elisabeth not only sealed the matrimonial bond, but also the business relationship, which led to the formation of the ­newly merged Browag AG. In 1935, Hermann was born in Paris (died 1987) and seven years later, along came his sister Marie (born 1942, died 2020). Owing to his commitment to the time-consuming management of the branch in Vienna, Paul did not move away from Vienna. Thus it came about that — some years later —Elisabeth and Paul separated. Paul decided to sell his company shares to Elisabeth, and to quit Browag AG. Their son Hermann decided to go to Vienna to study and to live with his grandparents, Hermann and Emma. Marie, on the other hand, wanted to study art in London and decided not to take up her mother’s offer of joining the company in Paris. Elisabeth Brown is considered one of the most successful female entrepreneurs of the 21st century and is leaving her entire fortune to charitable organizations with a focus on design.

19

Unstructured running text with highlights

In this version, the running text has already been enriched with a few support­ing symbols. Each symbol represents a certain type of information and is marked in yellow. Although at this stage we cannot yet talk about a visualization, a search, e.g. for Otto Wagner’s date of birth is made easier. The text is still unstructured.

▶ First name ◀ Family name ● Year of birth ◖ Year of death ■ Place of birth ◆Home town

20

The story of the entrepreneurial families

The success of the Browag AG company goes back to ▶ Hermann ◀ Wagner and ▶ James ◀ Brown. ▶ Hermann ◀ Wagner was born in ■ Vienna in ● 1871 (died ◖ 1961). During his student years he got to know ■ Vienna-born ▶ Emma Moser (1875), who was four years younger (died ◖ 1960). In 1900, ▶ Hermann and ▶ Emma married in ◆ Vienna, where the children ▶ Otto (● 1901, died ◖ 1924) and ▶ Paul (● 1914, died ◖ 2011) were also born. With the help of the ◀ Moser family’s financial resources, ▶ Hermann was able to set up Wagner Farben GmbH and establish it successfully in ◆ Vienna. After finishing school, ▶ Otto decided to go to ◆ Munich to study and ▶ Paul decided to follow in his father ▶ Hermann’s footsteps. ▶ Otto ◀ Wagner died tragically of tuberculosis in ◆ Munich at the age of 23. In the course of his business, ▶ Paul traveled to other countries and, in 1934, whilst in Paris, he met his future wife, ▶ Elisabeth ◀ Brown of Brown Chemicals Inc. The company Brown Chemicals Inc. had been founded in ■ London by ▶ James ◀ Brown (born ● 1882, died ◖ 1947). He benefited greatly from his marriage to ▶ Marie Durand (born ● 1879, died ◖ 1951), who came from a prosperous ■ Parisian family. ▶ James built up his business in ■ London as well as in ◆ Paris, where the couple spent most of their time. ▶ Marie and ▶ James had two daughters, ▶ Anna (born ● 1913, died ◖ 1996) and ▶ Elisabeth (born ● 1915, died ◖ 2014). ▶ Anna was born while the family spent some time in ■ London; she grew up there, went to a London boarding school, and spent the rest of her life in ◆ London. Her sister ▶ Elisabeth grew up in ■ Paris where she lived with her parents; from an early age, she joined papa ▶ James’s company, where she later met ▶ Paul ◀ Wagner. The wedding between ▶ Paul and ▶ Elisabeth not only sealed the matri­ monial bond, but also the business relationship, which led to the formation of the newly merged Browag AG. In ● 1935, ▶ Hermann was born in Paris (died ◖ 1987) and seven years later, along came his sister ▶ Marie (born ● 1942, died ◖ 2020). Owing to his commitment to the time-consuming management of the branch in Vienna, ▶ Paul did not move away from ◆ Vienna. Thus it came about that — some years later— ▶ Elisabeth and ▶ Paul sep­ arated. ▶ Paul decided to sell his company shares to ▶ Elisabeth, and to quit Browag AG. Their son ▶ Hermann decided to go to ◆ Vienna to study and to live with his grandparents, ▶ Hermann and ▶ Emma. ▶ Marie, on the other hand, wanted to study art in ◆ London and decided not to take up her mother’s offer of joining the company in ◆ Paris. ▶ Elisabeth ◀ Brown is considered one of the most successful female entrepreneurs of the 21 st century and is leaving her entire fortune to charitable organizations with a focus on design.

21

Running text as a list, structured with highlights

A mere compilation of data in the form of a list is not enough to structure the under­lying text. A list guides the eye through the listed items. The added ­value is not achieved until the sentence structure becomes comparable through identical arrangement. The symbols are used to make things even clearer. It is possible — albeit not yet very easy —  to detect a pattern in the lines.

▶ First name ◀ Family name ● Year of birth ◖ Year of death ■ Place of birth ◆Home town

22

The story of the entrepreneurial families

• ▶ Hermann ◀ Wagner, born ● 1871 in ◆ Vienna, lived in ◆ Vienna, died ◖ 1961, married to ▶ Emma ◀ Moser, children ▶ Otto and ▶ Paul • ▶ Emma ◀ Moser later ◀ Wagner, ● 1875 born in ■ Vienna, lived in ­◆ Vienna, died ◖ 1960, married to ▶ Hermann ◀ Wagner, children ▶ Otto and ▶ Paul • ▶ Otto ◀ Wagner, born ● 1901 in ■ Vienna, lived in ◆ Munich, died ◖ 1924 • ▶ Paul ◀ Wagner, born ● 1914 in ■ Vienna, lived in ◆ Vienna, died ◖ 2011, married to ▶ Elisabeth ◀ Brown, children ▶ Hermann and ▶ Marie • ▶ James ◀ Brown, born ● 1882 in ■ London, lived in ◆ Paris, died ◖ 1947, married to ▶ Marie ◀ Durand, children ▶ Elisabeth and ▶ Anna • ▶ Marie ◀ Durand later ◀ Brown, born ● 1879 in ■ Paris, lived in ◆ Paris, died ◖ 1951, children ▶ Elisabeth and ▶ Anna • ▶ Elisabeth ◀ Brown later ◀ Wagner, born ● 1915 in ■ Paris, lived in ◆ Paris, died ◖ 2014, married to ▶ Paul ◀ Wagner, children ▶ Hermann and ▶ Marie • ▶ Anna ◀ Brown, born ● 1913 in ■ London, lived in ◆ London, died ◖ 1996 • ▶ Hermann ◀ Wagner, born ● 1935 in ■ Paris, lived in ◆ Vienna, died ◖ 1987 • ▶ Marie ◀ Wagner, born ● 1942 in ■ Paris, lived in ◆ London, died ◖ 2020

23

Structure in table form

When, in a list consisting of sentences with an identical structure, regular ­spaces are inserted between the different parts of the sentence, a table is created. The placement of structured text elements in a grid generates good readability and comparability of the contents. The highlighting symbols used previously are no longer required here — the gap between the parts of the sentence orders the elements in accordance with their affiliation and columns are created.

▶ First name ◀ Family name ● Year of birth ◖ Year of death ■ Place of birth ◆Home town

24

The story of the entrepreneurial families

◀ Family name

▶ First name

■ Place of birth

● Year of birth

◆ Home town

◖ Year of death

▶ Children

◀ Wagner

▶ Hermann

■ Vienna

● 1871

◆ Vienna

◖ 1961

▶ Otto ▶ Paul

◀ Wagner (◀ Moser)

▶ Emma

■ Vienna

● 1875

◆ Vienna

◖ 1960

▶ Otto ▶ Paul

◀ Brown (◀ Durand)

▶ Marie

■ Paris

● 1879

◆ Paris

◖ 1951

▶ Anna ▶ Elisabeth

◀ Brown

▶ James

■ London

● 1882

◆ Paris

◖ 1947

▶ Anna ▶ Elisabeth

◀ Wagner

▶ Otto

■ Vienna

● 1901

◆ Munich

◖ 1924

◀ Brown

▶ Anna

■ London

● 1913

◆ London

◖ 1996

◀ Wagner

▶ Paul

■ Vienna

● 1914

◆ Vienna

◖ 2011

▶ Hermann ▶ Marie

◀ Wagner (◀ Brown)

▶ Elisabeth

■ Paris

● 1915

◆ Paris

◖ 2014

▶ Hermann ▶ Marie

◀ Wagner

▶ Hermann

■ Paris

● 1935

◆ Vienna

◖ 1987

◀ Wagner

▶ Marie

■ Paris

● 1942

◆ London

◖ 2020

25

Structure in table form, nested

Through a new arrangement, it is possible to organize a table such that repeating terms can be omitted. This has the effect of making patterns discernible already at the table stage. Nesting a table in this way increases the read­ability many times over. In this step it is always advisable to start with the column that has the fewest expressions. In this case it is the family name, with the two expressions Wagner and Brown.

▶ First name ◀ Family name ● Year of birth ◖ Year of death ■ Place of birth ◆Home town

26

The story of the entrepreneurial families

◀ Family name

■ Place of birth

◆ Home town

▶ First name

● Year of birth

◖ Year of death

▶ Children

◀ Wagner

■ Vienna

◆ Vienna

▶ Hermann

● 1871

◖ 1961

▶ Otto ▶ Paul

▶ Emma

● 1875

◖ 1960

▶ Otto ▶ Paul

▶ Paul

● 1914

◖ 2011

▶ Hermann ▶ Marie

◆ Munich

▶ Otto

● 1901

◖ 1924

■ Paris

◆ Vienna

▶ Hermann

● 1935

◖ 1987

■ Paris

◆ London

▶ Marie

● 1942

◖ 2020

■ Paris

◆ Paris

▶ Marie

● 1879

◖ 1951

▶ Anna ▶ Elisabeth

▶ Elisabeth

● 1915

◖ 2014

▶ Hermann ▶ Marie

◆ Paris

▶ James

● 1882

◖ 1947

▶ Anna ▶ Elisabeth

◆ London

▶ Anna

● 1913

◖ 1996

◀ Brown

■ London

27

Hierarchical structure

The construction of a family tree is founded on two basic elements: firstly, on a clear structure, as in a table, and secondly, on the further information about the connections between the various people mentioned in the text. The newly created order makes it possible to discern patterns even faster. Family trees are often used in visual­ ization to represent hierarchies in groups. In this concrete example, the family tree shows the sequence of relationships between the generations of the two families. Usually, a box represents ap ­ erson, and the nodes and linear connections represent the relationships ­between the members.

28

▶ First name ◀ Family name ● Year of birth ◖ Year of death ■ Place of birth ◆Home town

The story of the entrepreneurial families

◀ Wagner ▶ Hermann ■ Vienna ● 1871 ◖ 1961 ◆ Vienna

◀ Wagner (◀ Moser) ▶ Emma ■ Vienna ● 1875 ◖ 1960 ◆ Vienna

◀ Brown ▶ James ■ London ● 1882 ◖ 1947 ◆ Paris

◀ Brown (◀ Durand) ▶ Marie ■ Paris ● 1879 ◖ 1951 ◆ Paris

◀ Wagner ▶ Otto ■ Vienna ● 1901 ◖ 1924 ◆ Munich

◀ Wagner ▶ Paul ■ Vienna ● 1914 ◖ 2011 ◆ Vienna

◀ Wagner (◀ Brown) ▶ Elisabeth ■ Paris ● 1915 ◖ 2014 ◆ Paris

◀ Brown ▶ Anna ■ London ● 1913 ◖ 1996 ◆ London

◀ Wagner ▶ Hermann Jr. ■ Paris ● 1935 ◖ 1987 ◆ Vienna

◀ Wagner ▶ Marie Jr. ■ Paris ● 1942 ◖ 2020 ◆ London

29

1B From data to data cube In order to make it easier to orient the phases of data analysis and help with understanding the different ways a data set can be viewed, we have developed the data cube as a metaphor. We are not so much referring to the fact that each cube has exactly six equal sides, but instead are using this term in the metaphorical sense that a cube can be viewed from different sides. For this reason, we could also refer to the cube as a meta-cube or hyper-cube, but for simplicity’s sake we’ll just stick with data cube. For example, each column of a table, i.e. each data dimension, can be understood as one side of the cube. The steps proposed here for data analysis belong to the field of elementary exploratory statistics. We only want to introduce as much data science as is necessary for the next chapters to show different perspectives of the entrepreneurial families’ data set used as an example.

31

Data dimensions as perspectives on a data cube

Family name

First name

Place of birth Year of birth

Home town

Year of death

Wagner

Hermann

Vienna

1871

Vienna

1961

Wagner (Moser)

Emma

Vienna

1875

Vienna

1960

Brown (Durand)

Marie

Paris

1879

Paris

1951

Brown

James

London

1882

Paris

1947

Wagner

Otto

Vienna

1901

Munich

1924

Brown

Anna

London

1913

London

1996

Wagner

Paul

Vienna

1914

Vienna

2011

Wagner (Brown)

Elisabeth

Paris

1915

Paris

2014

Wagner

Hermann Jr.

Paris

1935

Vienna

1987

Wagner

Marie Jr.

Paris

1942

London

2020

32

The metaphor of a data cube helps us view a data set from a number of different angles. We refer to these perspectives as data ­dimensions (here, as columns of the table). The entrepreneurial family can be ­considered from the aspect of its family name, its home towns, or from any other perspective. It is important that the core always contains the same data set. Each viewing angle results in a different approach, analysis, statement, and story. The data from the first part of this chapter is listed here in the form of a table on the left-hand side. However, it is also possible to use the existing data to derive new data, groups, or categorizations: the first names can be used to infer the gender, the date of birth to infer the generation, the home town and the place of birth to infer relocations, the years of birth and death to infer the age at death, and from this, the age group too.This new data can be found in the table on the right-hand side.

Three of the eleven data dimensions as perspectives on three sides of the data cube of the entrepreneurial families.

■ Place of birth

● Year of birth

◀ Family name

Gender

Generation

Relocation

Age at death

Age group

M

1

N

90

> 85

F

1

N

85

70 – 85

F

1

N

72

70 – 85

M

1

Y

65

 85

M

3

Y

52

 85

F

1

N

85

70 – 85

F

1

N

72

70 – 85

M

1

Y

65

 85

M

3

Y

52

 85

F

1

N

85

70 – 85

F

1

N

72

70 – 85

M

1

Y

65

 85

M

3

Y

52

85 (3)

51

Area size 100% in an angle as a circle

2A.9 Quantity

Age group

When the total of all quantities adds up to 100 percent and each value is calculated as a percentage of a 360-­ degree circle, we obtain the wellknown pie chart. The fact that these are so widely used is due to com­ puter software, which makes it easy to design them. Constructing such a chart with a compass and ruler would be too labo­rious in everyday analog life.

70–85 (4)

< 70 (3)

> 85 (3)

52

Area size 100% in an angle as a circle segment

2A.10 Quantity

Age group

When there is not enough space for a full circle as shown on the left-hand side (2A.9) or it makes more sense to show the area as a segment, it is pos­sible to divide the 100 percent of values over 180 degrees or 90 degrees. Whether to choose a full circle or segment is more of a design decision than a mathematical one.

70–85 (4)

< 70 (3)

> 85 (3)

53

Area size 100% in an angle as a ring

2A.11 Quantity

Age group

Basically, this diagram is to be used and calculated in the same way as the pie chart 2A.9. This type of diagram is prefera­ble to a pie chart when additional information is to be placed in the center, e.g. the total quantity as a ­number. This diagram is often referred to as a donut diagram.

70–85 (4)

< 70 (3)

> 85 (3)

54

Area size 100% as a rectangle

2A.12 Quantity

Age group

The 100 percent does not have to be subdivided into segments of a circle, but can also be represented as a rectangle divided into smaller rect­ angles. This type of diagram is called a treemap. Calculating the different rectangle sizes is somewhat more complicated than calculating the size of slices in pie diagrams with a 360degree distribution.

< 70 (3) 70–85 (4) > 85 (3)

55

Area size 100% as a countable rectangle

2A.13 Quantity

Age group

The area of a rectangle can consist of 100 countable dots, thus representing 100 percent of a quantity distri­ bution. We have discovered this type of diagram as a combination of 2A.12 and 2A.6 and feel certain that its potential has not yet been fully exploited. Without the existence of computers, this type would probably be more popular because it can be manually produced, quickly and intuitively.

< 70 (3) 70–85 (4) > 85 (3)

56

2B Position

Points in a coordinate system are not negotiable; they have to be exactly right. Since these points represent positions in space, they are essential to the main message in these forms of presentation: each point has a very specific significance. We distinguish five position-related diagrammatical elements. Commonly, the most frequent forms of presentation are coordinate systems with two axes and geographical maps.

59

Points in space One coordinate axis

2B.1 Position

Home town

Points on a coordinate axis with a scale are a very precise method of repre­ senting data, because every value has a clearly defined position. One disadvantage of this is that the quantity cannot immediately be perceived as a value, but has to be deduced using its position. In addition, difficulties can arise when values overlap.

Persons 5 4

Vienna

3

Paris

2

London

1

Munich

0

60

Points in space Two coordinate axes

2B.2 Position

Home town

A typical coordinate system consists of two axes that open up space for placing values with two dimensions at an exact position. Compared to 2B.1, there is less potential for overlapping because the expressions (e.g. towns and number of people) are distributed on their own axis. Two-dimensional axes extending in both directions can also open up a negative number space, which creates a cross shape with four fields.

London Munich Paris Vienna

0

1

2

3

4

5

Persons

61

Points in space Three coordinate axes

2B.3 Position

Home town Relocation

Points in a three-dimensional space require three axes, which often im­ pinges on the readability; sometimes it is difficult to ascertain on which axis a point is located. In this case, the interactive possibilities of a turnable virtual space can provide a solution.

Persons 4 3 no

2 yes

1

Paris

Vienna

Munich

London

0

62

Points in space Polar axes

2B.4 Position

Home town

In a polar coordinate system, each point has a distinct position on a scale in connection with a direction. In this way, it is possible — as in 2B.2 — to avoid overlapping and to make further statements about a certain group of data.

Vienna Persons

Munich

Paris

London

63

Points in space Geographical coordinates

2B.5 Position

Home town

Maps can be understood as points in space with geographical coordinates. The respective points of the coordinates do not always have to be depicted in the form of an exact map, but can also be shown in a creative, mental grid that indicates directions and relationships. Countries do not always need to be shown with complex borders, but can also be depicted in a reduced form, depending on the intended statement.

London

Munich

Vienna

Paris

1 dot = 1 person

64

2C Relationship

The representation of relationships shows the connections between several points following a certain logic. Relationships can be represented as networks, hierarchically, and also nested. In all cases, the specific content determines the form of representation.

67

Network Non-hierarchical

2C.1 Relationship

Family

The network without a hierarchical structure (direction) is the basic diagram used to represent relationships. It is based on the principle that elements (nodes) can be connected to each other by edges (connections). A quantity can be represented using the number of connections. Furthermore, the length of the connecting line can be significant. Networks are used in a wide range of fields; how­ ever, sometimes — beyond a certain quantity of data — they are no longer clearly readable.

1 dot = 1 person

68

Network Hierarchical

2C.2 Relationship

Family

A hierarchical network is either nested or features a relationship sequence. The principle is very similar to that of 2C.1; the difference is, that here, there is a higher-level structure, possibly also a specific direction of reading. A family tree is an example of a type of diagram with a non-directional hier­ archy. When the lines indicate a direction (with an arrow), the hierarchy is shown to also indicate a process.

1 dot = 1 person

69

Network Linear

2C.3 Relationship

Family

A network presented in linear form contains all points on an axis, similar to a scale (2B.1). This opens up further possibilities for using the space differently and reordering the connections (nodes). Here too, it is possible to create applications with a fixed hierarchy or sequence. Overlaying lines indicate clusters.

1 dot = 1 person

70

Network Circular

2C.4 Relationship

Family

A network represented in a circle shows the relationships of a closed group. In this case the points (nodes) are placed on a circular line. Similar to the linear network (2C.3), clusters of connecting lines become apparent. In the case of a directional representation, it is also possible to represent two-way movements in the system.

1 dot = 1 person

71

Nesting Intersection sets

2C.5 Relationship

Family

Wagner

It is also possible to represent relationships without connecting lines in a defined space. In accordance with the basic rules of mathematical set theory, it is possible to form intersection sets, sub sets, and connection sets. In the example, the different levels and overlaps show specific information.

Brown

1 dot = 1 person

72

Nesting Stacking

2C.6 Relationship

Place of birth Family Paris W

Vienna

When relationships seem to indicate a hierarchical structure, but have no connecting lines (2C.2), stacking is a suitable form of representation; structures become clearly visible and have more detailed information at each level of this nesting.

London

W B

B

1 dot = 1 person

73

Nesting Matrix

2C.7 Relationship

Place of birth Family Paris

Vienna

A matrix is a table that shows twodimensional relationships. In this case, the focus is on the correlation between a certain column and a certain row. It is also possible to work with quantities at the intersection.

London

Wagner

Brown

1 dot = 1 person

74

3C Line 3B Shape 3A Color

3F Isotype 3E Contour 3D Pattern

3 Visual dimensions The 40 elements of the visual dimensions resolve the question as to how contents can be distinguished visually. We aim to demonstrate that, in addition to color distinction, there are many other options for translating a number or text in a visually attractive way that can be clearly understood. This chapter uses findings from the fields of graphic design, the psychology of perception, cartography, and pictorial statistics. The elements presented here are subdivided into the following six groups: color, shape, line, pattern, contour, and the special group of isotypes, according to Otto Neurath. We have analyzed these using several publications on pictorial statistics by Neurath; they are visual­ ized here for the first time in an abstract shape that is generally ­applicable (3F). The elements of the visual dimensions are particularly suitable for distinguishing the expressions of the “Category” data type. All those items of information that can be grouped together can be categorized, for example, family names as a category with the two expressions Wagner and Brown. Each individual element in this chap­ ter is symbolized by an icon; it is visualized using a basic example from our data set and supported by a short explanation.

77

3A Color

Color is one of the most frequently used means for expressing distinc­ tions in diagrams. Every software tool for data visualization uses color as standard for distinguishing data dimensions or value ranges. We are listing here six possible visual ele­ ments from the group of colors. Some principles are suitable for manual sketching whereas others are only useful in a digital context.

79

Basic colors

3A.1 Color

Family

Brown

Basic colors are easy to use and allow good differentiation. Having said that, there are only nine colors that  allow good differentiation: yellow, or­ ange, red, pink, violet, blue, and green, as well as brown and black. Color provides a good option for different­ iating categories when sketching visualizations.

Wagner

1 dot = 1 person

80

Saturation with white

3A.2 Color

Family

An alternative to using different colors is to mix a main color with white. In this way it is possible to create several different gradations from one color. It can also be aesthetically very pleas­ ing to work with fewer colors. This type of color gradation is particularly suit­ able for gradations in time sequences, for example light for newer and dark for older values.

Wagner

Brown

1 dot = 1 person

81

Saturation with black

3A.3 Color

Family

Saturation with black refers to the process of adding black to a color. Ex­ actly as in 3A.2, this saturation is well suited to representing sequences. Dark gradations are more likely to be seen as dirty or as obsolete points in time. Colors particularly suitable to gradation are shades of blue, green, and gray, because the human eye is trained to perceive the sky, a forest, and dusk.

Wagner

Brown

1 dot = 1 person

82

Color gradient

3A.4 Color

Family and birth name Durand– Brown

Brown

Color gradients can be used to distin­ guish data expressions. In this case, two or more colors are used to create a gradual change of color. The effect depends heavily on the contrast of the selected colors. This becomes partic­ ularly interesting when the color mix is associated with a meaning.

Wagner– Moser

Wagner

Brown– Wagner

1 dot = 1 person

83

Filled and empty

3A.5 Color

Family

Brown

A filled shape makes a strong con­­trast to an empty one. This can be per­ ceived as either full / empty or as much / little. The differentiation effect of this element can be multiplied by com­bining with different shapes and colors.

Wagner

1 dot = 1 person

84

Transparency

3A.6 Color

Family

Durand– Brown

The transparency of a shape does not become visible until several shapes are overlapped. It is primarily suitable for two-dimensional representations. Overlaying the shapes can denote additional meaning.

Wagner– Moser

Brown– Wagner

1 dot = 1 person

85

3B Shape

Besides color, form is one of the most important visual differentia­tors. Using forms in countable types of diagrams is particularly useful be­ cause they can be comprehended very quickly. Certain forms are also very suitable for a quick analysis using manual sketches. We have defined eight possible visual elements in the group of forms.

87

Geometrical

3B.1 Shape

Relocation

yes

There is a limited number of easily distinguishable shapes that also work visually in small formats. Shape as a distinguishing characteristic is also a very good tool for a quick hand sketch.

no

1 shape = 1 person

88

Organic

3B.2 Shape

Relocation

yes

Different organic shapes can be used to distinguish data expressions. When using two or more organic shapes, it is important to ensure that there is a clear enough distinction between them. Sometimes, abstract organic shapes can also be perceived as objects, for example, a flower. A deliberate selection of shape is therefore important.

no

1 shape = 1 person

89

Rotation

3B.3 Shape

Relocation

no

Rotation is particularly suitable for two expressions with the same shape. The expression can then be multi­ plied by combining it with color. The element of rotation can also be used to turn a graphic element, for example by 90 degrees, or to turn individual graphic elements around a point.

yes

1 shape = 1 person

90

Direction

3B.4 Shape

Relocation

no

When the characteristic of a shape is to point in a certain direction, it is possible to work deliberately with this element. For example, the triangle is a sufficiently dominant shape that can point in all four directions. yes

1 shape = 1 person

91

Stacking

3B.5 Shape

Relocation

By stacking shapes it is possible to visually distinguish two data expres­ sions from each other: differences in quantity and differences in shape. In the first case, the advantage is that the compact representation makes it possible to count quantities quickly whereas in the second case, a different visual form is created, as shown in the example on the left.

yes no

1 dot = 1 person

92

Blurring

3B.6 Shape

Relocation

Blurring a geometric shape works particularly well as a means of distin­ guishing when the same shape can also be seen with sharp edges. Blurring can be a good visual tool to indicate uncertainty in data. no

yes

1 dot = 1 person

93

Transformation

3B.7 Shape

Relocation

no

The transformation of a basic shape creates one, or several, new shapes. Transformation can be used to make a shape appear three-dimen­sionally in the space or to show ­movement. yes

1 shape = 1 person

94

Volume

3B.8 Shape

Relocation

yes

Each two-dimensional shape can also be represented three-dimensionally in the space. This makes it possible to represent volumes or quantities. It is important to keep an eye on the calcu­ lation to convert two-dimensional areas to three-dimensional bodies in order to avoid strong distortion in perception.

no

1 shape = 1 person

95

3C Line

Lines in their various forms are often underestimated as differentiators. Besides their use in diagrams that are based on the length of lines (2A), such as bar charts, different types of lines have great potential for rep­ resenting different relationships (2C). A line that connects two items can easily be enriched with additional information. It is possible to distin­ guish three to a maximum of five different types of lines. We are listing here eight different types of lines as visual dimensions.

97

Color

3C.1 Line

Age group

< 70

The line color is a frequently used and well-functioning distinguishing characteristic of categories in line lengths (2A) or relationship represen­ tations (2C). See also color (3A).

70–85

> 85

1 line = 1 person

98

Thickness

3C.2 Line

Age group

The thickness of a line is a clear dis­ tinguishing characteristic provided that there are not too many different thick­ nesses. Line thickness should not be used for quantities, but for categories, because different line lengths may easily lead to confusion. > 85 70–85

< 70

1 line = 1 person

99

Interruption

3C.3 Line

Age group

When dashed lines are used to make a distinction, it is important to ensure that the individual segments and line lengths are different enough. Dashed lines can be used for category data (different dashing) as well as for num­ ber data (dash length /gaps in relation to numbers).

70–85 (4) < 70 (3)

> 85 (3)

7.5 mm = 1 person

100

Roughing

3C.4 Line

Age group

A line can be modified to have a rounded or jagged wavy shape. Each wave frequency and shape can be assigned to a different characteristic expression of a category. It is also possible to count the number of peaks and troughs representing a unit.

< 70 (3)

> 85 (3) 70–85 (4)

7.5 mm = 1 person

101

Details

3C.5 Line

Age group

An often-used method in carto­ graphy is to enrich lines with various details. These may indicate roads, pathways, and railway tracks. When creating one’s own lines with details on maps, it is important to ensure that no wrong associations are created (for example, a state boundary). But it is also possible to use lines with details as a means of visual distinction, independent of maps.

< 70 (3)

70–85 (4)

> 85 (3)

7.5 mm = 1 person

102

Organic

3C.6 Line

Age group

< 70 (3)

By bending lines, numerous organic shapes can be created. This type of line can be used in support of a theme. In comparison with roughing (3C.4), the frequency of the wave is not ­relevant to the bending.

> 85 (3)

70–85 (4)

7.5 mm = 1 person

103

Detail Beginning or end

3C.7 Line

Age group

Different details at the beginning or end of a line offer a clear and simple way to establish distinguishability between data expressions. Here, it is important that these details are big enough to be easily distinguished.

< 70 (3) 70–85 (4) > 85 (3)

7.5 mm = 1 person

104

Arrow

3C.8 Line

Age group

Nowadays, it is no longer possible to imagine everyday life without arrows at the end of a line; everybody knows that this triangle at the end of a line indicates a direction. Arrows are important for processes and other directional diagrams and can also be used effectively in combination with non-directional lines.

< 70 (3) 70–85 (4) > 85 (3)

7.5 mm = 1 person

105

3D Pattern

Filling an area with a pattern can save time when filling an area manu­ ally and it can also save a little bit of ink in the printing process. Shapes filled with patterns should be large enough to allow clear differentiation. It is important to note that, depend­ ing on their size and density, patterns are perceived differently. On maps, patterns are often used to represent thematic levels. In our system we define five different elements of this group of patterns.

107

Texture

3D.1 Pattern

Generation

When using patterns to fill an area, it is best to choose a structure that is easily distinguishable. In general, this means that the elements have to be a certain size.

1st (4) 3rd (2) 2nd (4)

108

Broken lines

3D.2 Pattern

Generation

An area can be filled with lines of equal thickness and equal spacing as ap ­ attern. The gaps in the line can be used here as a distinguishing char­ acteristic. This principle can also be used to r­ epresent uncertainties in a data set.

3rd (2) 1st (4)

2nd (4)

109

Line thickness

3D.3 Pattern

Generation

Line thickness in a pattern only works well as an element when there are not too many data expressions. Similar to 3D.4, this element can be used for the visualization of density. Color can be used for additional support.

3rd (2) 1st (4)

2nd (4)

110

Density

3D.4 Pattern

Generation

In contrast to the line pattern used to fill an area, the pattern here does not consist of a line but of dots or other shapes. It is important that the basic pattern element is the same for all lev­ els of density so that it is possible to perceive the difference between them.

3rd (2) 1st (4)

2nd (4)

111

Direction

3D.5 Pattern

Generation

The direction of a pattern is a relatively weak differentiation. The element can be created by mirroring the pattern without changing the shape in any other way. For example, this principle can be successfully used for large areas in maps.

1st (4)

3rd (2)

2nd (4)

112

3E Contour

Contours can be used to add addi­ tional meaning to forms or to hold to­ gether groups of forms. Contours are also sometimes referred to as peri­ meter lines. As is the case with differ­ ent types of lines, it is important to note that not more than three to five different types of contours can be distinguished. We have identified five different elements.

115

Color

3E.1 Contour

Place of birth

London (2)

The contour of a shape combined with color is a very good distinguishing characteristic. Colored contours are a very quick method of producing small visualizations by hand. Likewise, filled two-dimensional shapes can be given additional meaning by using different contour colors. Paris (4)

Vienna (4)

116

Interruption

3E.2 Contour

Place of birth

London (2)

When using interruption in a contour it is important to ensure that there is an adequate difference in the gaps and line lengths, just as with the interrup­ tion of lines in 3C.3. This particularly applies to small sizes.

Paris (4)

Vienna (4)

117

Thickness

3E.3 Contour

Place of birth

The thickness of the contour refers to the thickness of the line. Depending on the size of the basic shape, not too many line thicknesses should be used because it is hard to distinguish them without magnification when the dis­ tance between the shapes is too small.

London (2)

Paris (4)

Vienna (4)

118

Shape

3E.4 Contour

Place of birth

The contour can have its own shape and stand alone or form the edge of an existing shape. Nesting a shape with a contour creates a new shape that can easily be distinguished.

London (2)

Paris (4)

Vienna (4)

119

Details

3E.5 Contour

Place of birth

Details can be combined well with a contour. This makes it possible to create new and easily distinguishable contours. See also 3C.5.

Vienna (4) Paris (4)

London (2)

120

3F Isotype

At the beginning of the 20 th century, Otto Neurath was an important Austrian scientist and pioneer of the so-called “Vienna Method of Pictorial Statistics”. Together with graphic artist Gerd Arntz, he devel­ oped the International System of Typographic Picture Education (Isotype), some of which is still in use today. Because Neurath has been a big source of inspiration for our data visualizations over the last 15 years, we want to dedicate the follow­ ing chapter to Isotype. Following our visual analysis of numerous publications, particularly, the Gesellschafts- und Wirtschaftsatlas from 1931, we have defined eight principles in eight elements. In contrast to the very abstract shapes in the previous chapters, the more figurative shapes used in the Isotype system are its special characteristic. We have, however, also worked out a few principles in our analysis that will also work for abstract shapes. We have already pointed out that figurative shapes trigger direct associations in the viewer and that, sometimes, additional and in­ appropriate meanings are read into them. There is a big risk of manip­ ulation or that misunderstandings arise in the context of an unintend­ ed cultural context. We therefore advise not to work with figurative shapes when the correct icons are not available or when an incorrect interpretation could, and is likely to, occur. The icons on the following pages have been taken from the visual Isotype vocabulary used in the nineteen-thirties (except for one icon representing women). We are conscious of the fact that these historic images do not meet current requirements ­regarding gender equality and inclusivity.

123

Figurative shape

3F.1 Isotype

Gender

Stylized figures in the Isotype system are a pictorial translation of concrete  terms (e.g. worker, chair). Nowadays, the terms “icon” and “pictogram” are more commonly used. With figura­ tive shapes it is important that they are shown to be of equal size and that they are countable. When one icon is larger than the other one, there is a risk of misinterpretation; in this case, the larger figure could be interpreted as an adult and the smaller one as a child.

female

male

124

Figurative shape Detail

3F.2 Isotype

Gender

When using figurative shapes with details, the basic shape should be large enough to make it easy to distin­ guish the variations. The smaller the icons are, the harder it is to recognize the details. The detail does not have to be within the shape but can also be placed next to the shape as a distin­ guishing characteristic. It is important to think clearly about how the new icon combinations will be interpreted.

female

male

125

Direction

3F.3 Isotype

Relocation

The direction in figurative shapes is suitable for terms that describe a movement. Here, it is important to en­ sure that the direction can be inferred, for example, from the position of the legs. Furthermore, when showing half of a figure or figurative shape, it is common sense to only make the split vertical.

yes

no

126

Background Shaded

3F.4 Isotype

Relocation

Shadows are suitable for both figura­ tive and abstract shapes. In the Iso­ type pictorial statistics they are used to visually translate multiples of a quanti­ ty. Nevertheless, the effect of depth created by shadow can also be used for other subjects.

yes

no

127

Background / Foreground

3F.5 Isotype

Relocation

Simple abstract details behind or in front of a shape are simple graphic devices for emphasizing or highlighting a certain statement. This element can easily be combined with figurative shapes as well as with abstract shapes.

yes

no

128

Background Area

3F.6 Isotype

Place of birth

A popular Isotype style device is to hold several, often countable, shapes together by means of a background area. This can be an abstract design, in one color, or with color gradients, or even the flag of a country. It is also possible to reverse the colors from the foreground shape to the background area.

yes

no

129

Frame or separating line

3F.7 Isotype

Relocation

A closed line can be used as a frame that holds a group of elements to­ gether. When the line is not closed but only separates an area it is called a separating line. In contrast to 3E, here, the line is not used to delineate the shape.

Vienna London Paris

130

Background Figurative

3F.8 Isotype

Place of birth

Besides a simple area (3F.6) and a simple frame (3F.7), an illustrative background is the last element of the Isotype principles for distinguish­ ing groups. Illustrations can provide a context for the information shown ­in the foreground. However, it is important to ensure that the background of the overall composition is not more visually dominant than the foreground.

Vienna

Paris

London

131

4B Grouping 4A Sorting

4 Structuring dimensions In this chapter the diagrammatical and visual elements are combined in a new graphic form with 15 structuring elements. This is where the magic of design really happens. The arrangement in the object (4A Sorting) and in the space (4B Grouping) defines the formative shape of the visualization. Sorting means defining the data dimension used for ordering a visualization. At the same time, this also determines the shape of the diagram. In order to avoid one single diagram containing too much data, the diagram can be subdivided into several diagrams, a process called grouping. Which type of sorting is best for readability and which serves best for the purpose of visualization can only be found out by trying and comparing. Each individual element in this chapter is symbolized by an icon; it is visualized using a basic example from our data set and supported by a short explanation.

133

4A Structuring by sorting Sorting means making a conscious decision as to which data dimension should be used for ordering a visualization. This order can then appear in linear or other shapes. In this way, different visual representation shapes are created with the same data. Sorting the data places a certain focus on the data set and thus controls the story it is meant to tell. For example, in our data set the age, the gender, but also the place of birth, can move to the foreground. All data that does not have a unique position can be sorted (see the position-related elements 2B). Texts are always sorted alphabetically and numbers hierarchically. The bar chart in the following icons are representative of all dia­ grammatical elements and can be replaced by them. (See also multidimensional visualization examples.) We have defined nine possible ways of sorting, which are explained and visualized on the next pages. In chapter five we explain how sorting functions in practice.

135

Linear axis

4A.1 Sorting

Age

Depending on the intended statement and the space available, it is also possible to arrange the lines turned by 90 degrees.

Otto (23) Hermann Jr. (52) James (65) Marie (72) Marie Jr. (78) Anna (83)

Example: all ten persons are sorted by age (ascending) on a linear axis (turned by 90 degrees).

Emma (85) Hermann (90) Paul (97) Elisabeth (99)

1 dot = 10 years

136

Linear axis with line break

4A.2 Sorting

Sorting characteristics on a linear axis with a line break may be used when only limited space is available. It is important to continue the sorting after the line break along the line.

(7 ie ar

M

Ja

2)

5)

(6 es

m

m

78 ie

Jr .(

3) (8 M

ar

a nn

m

Example: all ten persons are sorted by age (ascending) along a line with a line break.

et ab

is El

Pa

ul

(9

7)

h

(9

9)

A

H

Em

er

m

a

an

(8

n

5)

(9

)

0)

H

O

tt

er

o

(2

an

3)

n

Jr .(

52

)

Age

137

Circular Parallel

4A.3 Sorting

Age group

Using parallel circular lines as a means of sorting is better suited to the comparison of data in percent than for the representation of absolute numbers. This method is only suitable for quantities, not for two-dimensional areas.

< 70 (3) 70–85 (4) > 85 (3)

Example: all ten persons are sorted by three age categories (ascending) using parallel circular lines.

60° = 1 person

138

4A.4 Sorting

9)

Radially at an angle

With radial sorting always using the same angle it is possible to produce star-shaped diagrams. When twodimensional areas are used instead of lines there may be overlaps.

El

isa

be

th

(9

Age

P

l au

(9

7)

rm He

(90

ann

E m m a (8 A n n a (8 3

M Ja

He

ie

es

rie

(7

Jr. (

5)

)

Example: all ten persons sorted by age (descending) in radial formation with the same angle.

78)

2)

(6

Jr. (

5)

ann

52)

2.5 mm = 10 years

ar

m

)

rm

O tto (2 3

Ma

)

139

Radially Evenly distributed

Pa

ul

(9

7)

Here, the difference to 4A.4 is that the values are distributed using the entire 360 degrees of the circle. To make a precise comparison of the possible values, it may be helpful to show a scale in the background.

Ot

m Her

65)

Emm

(7

Example: all ten persons sorted radially by age (descending) with even ­distribution.

5)

3)

ie

)

(8

ar

a (8

(90

na

M

ann

An

Marie Jr. (78)

2)

es (

(52)

3)

Jam

n J r.

(2

man

to

Her

Elisabeth (99)

Age

4A.5 Sorting

2.5 mm = 10 years

140

Radially Evenly distributed on shape

sab eth

)

3) O tto (2

(97

)

ul

(99

Pa

Starting the evenly distributed sorting lines from a circle or other geomet­ rical shapes creates space. In this way it would also be possible to represent two-dimensional areas or even, within the circle, negative values.

Eli

Age

4A.6 Sorting

H

m er

an

n

.( Jr

Ja m e s (6 H e rm a n

52

)

5)

n (9 0 )

Ma 5)

Example: all ten persons sorted radially by age (ascending) with even distribution, starting from a circle.

)

rie Jr. (78 )

(8 3 )

1 dot = 10 years

(72

Ma

m

(8

Anna

Em

a

rie

141

4A.7 Sorting

(52)

Axis Free shape

65

)

n J r.

The line of an axis does not always have to be straight. Sorted values can also be arranged along a curved line.

s(

man

(7

2)

me

Her

M

ar

ie

Ja

O tto (2 3)

Age

M

a

J rie

An

7 r. (

na

8)

(83

Emma

)

Example: all ten persons sorted by age (ascending) along a curved line.

(8 5 )

Her ma nn (90 ) Pa u l

Eli 2.5 mm = 10 years

sa

be

(97)

th

(99

)

142

Linear axis Stacked

4A.8 Sorting

Home town Gender

Stacked sorting requires two data dimensions. In stacked diagrams (e.g. stacked bar charts) the values must be sorted and grouped at the same time.

Munich London Paris

Example: all ten persons sorted by home town and stacked by gender.

Vienna

1 dot = 1 person / filled dot = female

143

Free without axis or point

There are visualizations without an axis or any sorting. Here, the perception principle of proximity and distance applies, i.e. close elements are considered to be related.

A

nn

a

(8

3)

Age

4A.9 Sorting

H m an n

ul

7)

Jr .( ie

El

is

M

ab

ar

et

h

3)

(9

(2

9)

o

78

tt

0)

O

(9

(9

)

er

Pa

H 2) (7 M

ar

ie

5) (8 a m

)

Em

52

5)

.( Jr

(6

n

es

an

m

m

Ja

er

Example: all ten persons arranged randomly in the space.

144

4B Structuring by grouping Grouping is the second structure-giving dimension. In this process, a multidimensional diagram is subdivided in accordance with a data dimension and arranged in the space. The two principles of grouping are repetition and juxtaposition. The advantage of grouping is based on the way humans recognize patterns: our eyes are trained to visually perceive even very small differences. Groups can be placed in the available space in a grid or using some other principle. In the following representations of elements all groups are re­presented with identical sorting based on a linear axis (4A.1). The elements of 4A.1 can be replaced by all elements in group 4A. We have defined six possible principles of grouping.

147

Linear

4B.1 Grouping

Home town Age

Linear grouping is generally only suitable when there are not many groups because the space in a row is often very limited. It is also possible to turn the arrangement by 90 degrees.

Munich

London

Paris

Example: all ten persons sorted by age (ascending) and grouped by their four home towns along a line.

Vienna

1 dot = 10 years

148

Grid

4B.2 Grouping

Home town Age

When groups are arranged in a grid, the terms “Small multiples” or “Facetting” are also used. With this principle it is possible to use the space in a very efficient and structured way.

Munich

London

Example: all ten persons sorted by age (ascending) and grouped by their four home towns in the grid. Paris

Vienna

2.5 mm = 10 years

149

On shape or angle

4B.3 Grouping

on

ic un

nd

Example: all ten persons sorted by age (ascending) and grouped by their four home towns around a geometric shape.

V

ris

ie

Pa

nn

a

M

When groups are arranged around a circle or a different geometric shape one needs to note that the radius ­increases as the number of groups increases.

Lo

h

Home town Age

150

Mirroring

4B.4 Grouping

Gender Age

When using mirroring for grouping it is important to ensure that the data points opposite each other are com­ parable. This principle only works when there are opposites, i.e. between two groups.

Otto (23) Marie (72) Hermann Jr. (52) Marie Jr. (78) James (65) Anna (83)

Example: all ten persons sorted by age and grouped mirrored by gender.

Hermann (90) Emma (85) Paul (97) Elisabeth (99)

2.5 mm = 10 years

151

Free in space

4B.5 Grouping

Home town Age

Mun

When groups are placed freely in space, the available format can be fully utilized. The arrangement of groups can be random or follow certain design criteria. ich

Lond

on

Example: all ten persons sorted by age (ascending) and grouped by their four home towns freely in space. Pa r i

s

V ie n

na

2.5 mm = 10 years

152

On levels

4B.6 Grouping

Gender Age

He

He Pa

ul

Jam rm

(97

an

n(

90

rm

es

an

(65

Em be

th

(99

r. ( 5

(23

When groups are placed on levels in space it is important to note that overlaps may make it difficult to read the diagram.

)

2)

)

)

An sa

nJ

to

)

Ma

Eli

Ot

)

ma

(85

na

Ma rie

(83

Jr.

(78

rie

(72

Example: all ten persons sorted by age (descending) and grouped by their gender on two levels.

)

)

)

) 1 dot = 10 years

153

5 The variety of ­multidimensional ­visualizations Using 26 innovative diagram variations, this chapter shows the connection and concrete application of all the elements and data described in the previous chapters. This is just a small selection of the numerous combinations that are possible with the ­Modular Information Design (MID) system for the visualization of complexity. All 80 elements are used at least once. In each visualization, a listing of the elements and data dimensions used indicates, in a step-bystep fashion, how this is structured. The examples show that even with just a few data dimensions and different elements, it is possible to achieve great design variety. Hopefully this will motivate the reader to try to design new forms of representation because the necessary attention relies only on attractive and easy-to-understand visualizations. To design is to think. That means trying different combinations in order to form variations. Mapping (see instructions on p. 160) selected elements on the selected data dimensions makes it easier to understand the data set and it is possible to discover stories that could not be detected from a simple table. Creating variations repeatedly not only helps to view the existing data from a different perspective, but also enables us to “think outside of the box” when trying to solve a problem.

154

By varying selected data dimensions and elements in a deliberate way, it is possible to create many different and new visualizations in a quick and focused manner. We need to design and compare several variations in order to be able to assess which is the most suitable for the selected data dimensions and which is the most readable. Sometimes a small subject-related question helps to make a decision. For example: In year X, which country had the largest proportion Y of women in group Z? Multidimensional visualizations cannot be created using just one correct software tool. It is always worth starting with pen and paper. The next step may involve graphics software, an online tool, or just a code. A selection of our favorites is included in the Appendix. However, there is no such thing as the perfect tool.

155

Overview of all 80 elements 2A Quantity

Diagrammatical dimensions

2B Position

2C Relationship

Visual dimensions

3A Color

3B Shape

156

3C Line

3D Pattern

3E Contour

3F Isotype

Structuring dimensions

4A Sorting

4B Grouping

157

How the MID system works

The MID system maps the elements of the diagrammatical, visual, and structuring dimensions onto the data dimensions of the data cube. By mapping the elements on the data dimensions, a multidimensional visualization is created. This always requires, as a minimum, a diagrammatical (blue), a visual (red), and a structuring (green) element. On principle, only one element is mapped on to the data cube for each perspective, i.e. the respective data dimension. An exception exists regarding the elements of the structuring dimension of sorting (4A). These are always tied to a blue or red element.

158

On the one hand, MID can be used to encode new forms of representation, i.e. diagram types; on the other hand, MID can also be used to decode the structure of existing visualizations. The example on the righthand side shows the mapping of five elements on four data dimensions. Detailed instructions can be found on the following page.

Mapping (Encoding)

Element of diagrammatic dimensions

Element of visual dimensions

Element of structuring dimensions

Data dimensions

Decoding

Multidimensional variants

159

Mapping four data ­dimensions — instructions

Age A

D Gender C

Family B

Generation E The instructions show these selected elements for a concrete visual­ization. By replacing ­ elements it is possible to create a large diversity of ­variations. Here, four data dimensions (yellow) and five elements of the diagrammatical, visual, and structuring dimensions are combined. (A–E, blue, red, green). Sorting (D) never has its own data dimension, but always in­volves a connection with another element.

160

A

B

2 mm = 10 years Wagner Brown Female Male

Age → 10 expressions

Family → 2 expressions

C

D

Gender → 2 expressions

Sorting by age, ascending

E

Every visualization can be built up step by step. The sequence in which this is carried out is less important than actually carrying out the steps. (The example only shows one of many possible allocations of the elements to the data dimensions.) During the process of sketching, several steps are often carried out simultaneously. However, it is worth retaining individual steps, for example the line length, to be used at a later date.

Grouping by generation → 3 expressions

161

How sorting works

For these visualizations, the same three data dimensions (age, family, and gender), as well as the same elements of the system (2A.1, 3A.2, 3C.1, 4A.5), were used four times. Therefore, the form of sorting (element 4A.5) is the same, only the organization of the data differs: in each case a different data dimension is in the foreground and this determines the sequence in which the ten data points are shown. This means that the star always looks a little dif­ferent, and the story always unfolds a little differently. In one case the age is in focus, in another case the family, and in two other cases, the gender.

All ten family members are radially sorted with even distribution ­according to the data dimension selected. A (Age)

The sorting by age clearly indicates that the three people who died at the youngest age are men.

B (Family)

The sorting by family highlights the fact that the three members of the Brown family are closer together in age than the members of the Wagner family.

C (Gender, Age)

The sorting by gender indicates that the difference in the age of the women is much smaller than that of the men.

D (Gender, Family, Age)

This sorting helps to compare the age, not only within the gender, but also — at the same time— within the families. When sorting is by quantity (here, age), the orga­nization is fixed. When sorting a data dimension, for example, the category of gender, it is mandatory to define whether to sort by family or by age first.

162

A

B

2 mm = 10 years Wagner Brown Female Male

Sorted by 1. Age

Sorted by 1. Family 2. Age

C

D

Sorted by 1. Gender 2. Age

Sorted by 1. Gender, 2. Family 3. Age

163

Rays Age as line length, sorted radially at an angle and grouped freely in space by generation.

James (65)

Marie (72)

Emma (85) Hermann (90)

Anna (83)

First generation

Elisabeth (99)

Marie Jr. (78)

Hermann Jr. (52)

Paul (97)

Otto (23)

Second generation

Third generation

In each generation the number of women and men is in balance. 164

Example 1

Age 2A.1

6.25 mm = 10 years

Family

Gender

Home town

3C.1

3C.3

3C.7

Wagner Brown

Female Male

London Munich Paris

Generation 4A.4

4B.5

Sorted by age

Grouped by generation

Elements used 2A.1 3C.1 3C.3 3C.7 4A.4 4B.5

Quantity. Line length Line. Color Line. Interruption Line. Detail. Beginning or end Sorting. Radially at an angle Grouping. Free in space 165

Vienna

Semi-circles Age by line length, sorted parallel circular by age and grouped mirrored by relocation.

0

0

25

o

O

tt

25

Not relocated

Relocated

50

50

He rm a nn J

r. Ja

m

Elisabeth ar

ie

Pa u l

J r.

Her m a n n

75

100

Emm

a

An

na

rie

es

M

M

a

75

100

In the Wagner and Brown families, the men are more likely to relocate than the women. People live longer when they do not relocate. 166

Example 2

Age 2A.1

180° = 100 years

Family

Gender

Home town

3C.1

3C.3

3C.7

Wagner Brown

Female Male

London Munich Paris

Relocation 4A.3

4B.4

Sorted by age

Grouped by relocation

Elements used 2A.1 3C.1 3C.3 3C.7 4A.3 4B.4

Quantity. Line length Line. Color Line. Interruption Line. Detail. Beginning or end Sorting. Circular. Parallel Grouping. Mirroring 167

Vienna

Sunrise

es m Ja 5) (6

M (7 arie 8) Jr.

Ot to (23 )

He (52 rma nn )

J r.

Elis (99 abeth )

Pa u l (9 7 )

An na (83 )

an Herm (90)

rie Ma ) 2 7 (

ma Em ) (85

n

Age as countable line length, arranged in a radial layout with equal spacing, based on a free shape.

On average, the people  from the first and second ge­n­erations have lived longer than those from the third ­generation.

168

Example 3

Age 2A.2

1 dot = 10 years ½ dot = 5 years

Family

Gender

Generation

3A.1

3A.2

3C.6

Wagner Brown

Female Male

1st 2nd 3rd

4A.7

Sorted by generation and age

Elements used 2A.2 3A.1 3A.2 3C.6 4A.7

Quantity. Line length. Countable Color. Basic colors Color. Saturation with white Line. Organic Sorting. Axis. Free shape 169

Tower Age as line length with line break, sorted by age,  stacked on linear axis and grouped mirrored by relocation.

Otto (23) Hermann Jr. (52)

James (65) Marie Jr. (78) Relocated Not relocated

Marie (72)

Anna (83)

Emma (85)

Hermann (90)

Paul (97)

Elisabeth (99)

On average, the people who relocated lived to an older age than those who did not relocate. 170

Example 4

Age 2A.3

20 mm = 10 years 1 line = 25 years

Family

Gender

Home town

3C.1

3C.3

3C.7

Wagner Brown

Female Male

London Munich Paris

Relocation 4A.8

4B.4

Sorted by age

Grouped by relocation

Elements used 2A.3 Quantity. Line length with line break 3C.1 Line. Color 3C.3 Line. Interruption 3C.7 Line. Detail. Beginning or end 4A.8 Sorting. Linear axis. Stacked 4B.4 Grouping. Mirroring 171

Vienna

Domino Age as countable line length with line break, sorted by age on linear axis with line break.

Otto (23) Hermann Jr. (52)

James (65)

Marie (72)

Marie Jr. (78) Anna (83)

Emma (85)

Hermann (90)

Paul (97) Elisabeth (99)

The three people who died at the youngest ages were all men. 172

Example 5

Age 2A.4

1 dot = 5 years ½ dot = 2 ½ years

Family

Gender

Relocation

3A.1

3E.2

3E.3

Wagner Brown

Female Male

Yes No

4A.2

Sorted by age

Elements used 2A.4 Quantity. Line length. Countable with line break 3A.1 Color. Basic colors 3E.2 Contour. Interruption 3E.3 Contour. Thickness 4A.2 Sorting. Linear axis with line break 173

Firework Age as a countable area, sorted radially by age evenly distributed and arranged freely without axis or point. Generation as area.

Otto (23) James (65)

Marie (72)

Hermann Jr. (52)

Marie Jr. (78)

Emma (85)

Anna (83)

Hermann (90)

Paul (97)

Elisabeth (99)

There are more Wagners than Browns. 174

Example 6

Generation

Age

2A.5

2A.6

S = 1st M = 2nd L = 3rd

1 line = 5 years

Family

Gender

Relocation

3C.1

3C.7

3A.2 3F.5

Wagner Brown

Female Male

4A.4

4A.9

Age

Sorted free without axis or point

Yes No

Elements used 2A.5 2A.6 3C.1 3C.7 3F.5 4A.4 4A.9

Quantity. Area size Quantity. Area size. Countable Line. Color Line. Detail. Beginning or end Isotype. Background / Foreground Sorting. Radially at an angle Sorting. Free without axis or point 175

Swarm Age as countable area, sorted radially by age evenly ­distributed on a circle.

Otto (23)

Hermann Jr. (52)

Elisabeth (99)

Paul (97)

James (65)

Hermann (90)

Marie (72)

Emma (85)

Marie Jr. (78) Anna (83)

There are three pairs of people of a similar age, each belonging to the same generation, but of different gender. 176

Example 7

Age 2A.6

1 shape = 1 year

Generation

Family

Gender

3B.3

3C.1

3C.4

0° = 1st 120° = 2nd 240° = 3rd

Wagner Brown

Female Male

4A.6

Sorted by age

Elements used 2A.6 3B.3 3C.1 3C.4 4A.6

Quantity. Area size. Countable Shape. Rotation Line. Color Line. Roughing Sorting. Radially. Evenly distributed on shape 177

Windmill Age group as area, 100 percent as circle from the center, sorted radi­ ally, evenly distributed by age and grouped freely in space by home town.

Paris

Vienna Hermann Jr. (52)

James (65)

Marie (72)

Emma (85)

Paul (97)

Elisabeth (99)

Hermann (90)

Munich

London Otto (23) Marie Jr. (78)

Anna (83)

At each of the home towns there is one member of the Wagner family. 178

Example 8

Age group 2A.7

S  85

Family

Relocation

Gender

3A.1

3A.4

3D.5

Wagner Brown

Yes No

Female Male

Home town 4A.5

4B.2

Sorted by age

Grouped by home town

Elements used 2A.7 3A.1 3A.4 3D.5 4A.5 4B.2

Quantity. Area size 100% as a circle from the center out Color. Basic colors Color. Color gradient Pattern. Direction Sorting. Radially. Evenly distributed Grouping. Grid 179

Origami Age as area, 100 percent as a circle from the edge, sorted radially and evenly distrib­uted on a circle by home town and age, and grouped mirrored by relocation.

London

Otto (23)

Anna (83)

Munich

100 75 50

Marie Jr. (78)

25

Hermann Jr. (52)

0 25 50 75 100

Elisabeth (99)

Emma (85) Relocated

Marie (72)

Paris

Hermann (90)

Not relocated

James (65)

Paul (97)

Vienna

Members of the Wagner family only live in Munich and Vienna. 180

Example 9

Age 2A.8

36° = 1 person Angle = Age

Family

Gender

Home town

3A.1

3D.1

3F.6 3F.7

Wagner Brown

Female Male

Separated by home town

Relocation 4A.6

4B.4

Sorted by home town and age

Grouped by relocation

Elements used 2A.8 Quantity. Area size 100% as a circle from the edge in 3A.1 Color. Basic colors 3D.1 Pattern. Texture 3F.6 Isotype. Background. Area 4A.6 Sorting. Radially. Evenly distributed on shape 4B.4 Grouping. Mirroring 181

Pie Age as area 100 percent in the angle as circle, sorted radially, evenly distributed by age, and grouped linearly by place of birth.

London

Vienna

Otto (23)

Anna (83)

Emma (85)

James (65)

Hermann (90)

Elisabeth (99)

Paul (97)

Marie Jr. (78)

Paris

Marie (72)

Hermann Jr. (52)

The sum of the age of all people born in Vienna is very similar to that of the people born in Paris. 182

Example 10

Age 2A.9

4.8° = 10 years

Family

Generation

Relocation

Gender

3A.1

3A.3

3A.4

3D.4

Wagner Brown

1st 2nd 3rd

Yes No

Place of birth 4A.4

4B.1

Sorted by age

Grouped by place of birth

Elements used 2A.9 Quantity. Area size 100% in an angle as a circle 3A.1 Color. Basic colors 3A.3 Color. Saturation with black 3A.4 Color. Color gradient 3D.4 Pattern. Density 4A.4 Sorting. Radially at an angle 4B.1 Grouping. Linear 183

Female Male

Fan Age as two-dimensional area 100 percent in the angle as a circle segment, sorted radially in the angle by family and gener­ation, grouped by place of birth.

Emma (85) Hermann (90) Otto (23) Paul (97)

Anna (83) Vienna James (65)

London

Paris

Elisabeth (99) Marie (72)

Marie Jr. (78)

Hermann Jr. (52)

The first two generations of the Wagner family ­account for half of all the people and most of them belong to the oldest group. 184

Example 11

Age 2A.10

2.4° = 10 years

Family

Generation

Gender

3A.1

3A.3

3D.4

Wagner Brown

1st 2nd 3rd

Female Male

Place of birth 4A.4

4B.3

Sorted by family and generation

Grouped by place of birth

Elements used 2A.10 Quantity. Area size 100% in an angle as a circle segment 3A.1 Color. Basic colors 3A.3 Color. Saturation with black 3D.4 Pattern. Density 4A.4 Sorting. Radially at an angle 4B.3 Grouping. On shape or angle 185

Ring Age as area 100 percent in the angle as ring, sorted by age.

Otto (23)

Elisabeth (99)

Hermann Jr. (52)

James (65)

Paul (97)

Marie (72)

Hermann (90)

Marie Jr. (78)

Emma (85)

Anna (83)

The Browns only feature in the first two generations. 186

Example 12

Age

Age

2A.11

2A.13

4.8° = 10 Jahre

1 line = 3 years

Family

Generation

Gender

3E.1

3E.4

3F.2

Wagner Brown

1st 2nd 3rd

Female Male

4A.6

Sorted by age

Elements used 2A.11 Quantity. Area size 100% in an angle as a ring 2A.13 Quantity. Area size 100% as a countable rectangle 3E.1 Contour. Color 3E.4 Contour. Shape 3F.2 Isotype. Figurative Shape. Detail 4A.6 Sorting. Radially. Evenly distributed on shape 187

Parcels Age as area 100 percent distributed by percentage.

Otto (23)

Hermann (90)

Paul (97)

James (65)

Hermann Jr. (52)

Marie (72)

Marie Jr. (78)

Emma (85)

Elisabeth (99)

Anna (83)

Those who relocated had a shorter life than those who did not relocate. 188

Example 13

Age 2A.12

Total area = 110 × 110 mm = 744 years

Family

Relocation

Gender

First name

3A.1

3B.8

3D.1

3F.7

Wagner Brown

Yes No

Female Male

4A.1

Sorted by family

Elements used 2A.12 Quantity. Area size 100% as a rectangle 3A.1 Color. Basic colors 3B.8 Shape. Volume 3D.1 Pattern. Texture 3F.7 Isotype. Frame or separating line 4A.1 Sorting. Linear axis 189

Separated by person

Display Age as countable area 100 percent, sorted linearly by family.

Otto (23) James (65)

Hermann Jr. (52)

Marie (72)

Hermann (90)

Paul (97)

Marie Jr. (78)

Elisabeth (99)

Anna (83) Emma (85)

The number of people in the Brown family account for 30 percent and their total age for 40 percent. 190

Example 14

Age 2A.13

1 dot = 7.5 years

Family

Generation

Gender

First name

3A.1

3A.4

3B.2

3F.7

Wagner Brown

1st 2nd 3rd

Female Male

4A.1

Sorted by family

Elements used 2A.13 Quantity. Area size 100% as a countable rectangle 3A.1 Color. Basic colors 3A.4 Color. Color gradient 3B.2 Shape. Organic 3F.7 Isotype. Frame or separating line 4A.1 Sorting. Linear axis 191

Separated by person

Pyramid Positioned by birth year on a coordinate axis. Age as line length.

1950

Marie Jr. (78) 1940

Third generation

Hermann Jr. (52) 1930

1920

Elisabeth (99) Paul (97) Anna (83) 1910

Second generation Otto (23)

1900

1890

James (65) Marie (72)

1880

First generation

Emma (85) Hermann (90)

1870

The people in the first generation are closest to each other in age (of death). 192

Example 15

Age

Year of birth

2A.1

2B.1

6.25 mm = 10 years

Y = Decades

Family

Gender

Generation

3C.1

3C.3

3F.6

Wagner Brown

Female Male

Separated by generation

Elements used 2A.1 2B.1 3C.1 3C.3 3F.6

Quantity. Line length Position. Points in space. One coordinate axis Line. Color Line. Interruption Isotype. Background. Area 193

Layers Positioned by age and year of birth /death between two coordinate axes. Age as line length.

Elisabeth (99) Paul (97)

100

Hermann (90)

90

Emma (85) Anna (83) 80

Marie Jr. (78) Marie (72)

70

James (65) 60

First generation Hermann Jr. (52)

50

Third generation 40

30

Otto (23)

20

Second generation 10

0 1870

1880 1890 1900

1910

1920

1930 1940 1950 1960

1970

1980 1990 2000 2010 2020

The person who died first and the two people who lived the longest all belong to the second generation. 194

Example 16

Year of birth Year of death 2B.2

X = Decades Y = Age

Family

Gender

Generation

3C.1

3C.3

3F.7

Wagner Brown

Female Male

Separated by generation

Elements used 2B.2 Position. Points in space. Two coordinate axes 3C.1 Line. Color 3C.3 Line. Interruption 3F.7 Isotype. Frame or separating line 195

Flags Age as line length, year of birth /death and relocation positioned between three coordinate axes and sorted by relocation and age.

2020

2000

1980

1960

1940

1920

1900

1880

Relocated 1860

Marie (72)

Anna (83)

Emma Hermann Paul (85) (90) (97)

Elisabeth (99)

Otto Hermann Jr. James (23) (52) (65)

Marie Jr. (78)

Not relocated

Nobody from the Brown family has lived in the 21 st century. 196

Example 17

Year of birth Year of death Relocation 2B.3

X = Person Y = Birth / death Z = Relocation

Family

Relocation

Gender

3A.1

3A.4

3D.1

Wagner Brown

Yes No

Female Male

4A.1

Sorted by relocation and age

Elements used 2B.3 Position. Points in space. Three coordinate axes 3A.1 Color. Basic colors 3A.4 Color. Color gradient 3D.1 Pattern. Texture 4A.1 Sorting. Linear axis 197

Target Age as line length, ­p ositioned on polar axes by year of birth /death and sorted by age.

Elisabeth (99)

2020

Otto (23)

2000 1980

Paul (97)

Hermann Jr. (52)

1960 1940 1920 1900 1880

Hermann (90)

James (65)

Emma (85)

Marie (72)

Anna (83)

Marie Jr. (78)

The four people with the longest life are members of the Wagner family. 198

Example 18

Year of birth Year of death 2B.4

Axis = Decades Sector = Person

Family

Gender

Relocation

3A.1

3D.1

3E.3

Wagner Brown

Female Male

Yes No

4A.5

Sorted by age

Elements used 2B.4 3A.1 3D.1 3E.3 4A.5

Position. Points in space. Polar axes Color. Basic colors Pattern. Texture Contour. Thickness Sorting. Radially. Evenly distributed 199

Movement chart

Paris

Vienna

Paris has the most relocations. 200

Paul

Hermann

Emma

Munich

Hermann Jr.

Otto

Elisabeth

London

Marie

James

Anna

Marie Jr.

Place of birth and home town positioned as geographic coordinates and set in relation to each other as a network without hierarchy.

Example 19

Home town Place of birth

Relocation

2B.5

2C.1

Family

Home town Place of birth

Relocation

Gender

3A.1

3A.5

3C.8

3F.2

Wagner Brown

Home town Place of birth

Yes

Elements used 2B.5 Position. Points in space. Geographical coordinates 2C.1 Relationship. Network. Non-hierarchical 3A.1 Color. Basic colors 3A.5 Color. Filled and empty 3C.8 Line. Arrow 3F.2 Isotype. Figurative Shape. Detail 201

Female Male

Lollipop Home towns and family members set in relation to each other as a network without hierarchy.

Emma

Hermann

Hermann Jr. Paul

Elisabeth

Marie

Otto

James

Anna

Marie Jr.

Overall, four people relo­cated; the majority of these were male members of the Wagner family. 202

Example 20

Home town 2C.1

Family

Home town

Relocation

Gender

3A.1

3B.1

3B.7

3D.2

Wagner Brown

London Munich Paris

Yes No

Vienna

Home town 4A.4

4B.3

Sorted by family

Grouped by home town

Elements used 2C.1 3A.1 3B.1 3B.7 3D.2 4A.4 4B.3

Relationship. Network. Non-hierarchical Color. Basic colors Shape. Geometrical Shape. Transformation Pattern. Broken lines Sorting. Radially at an angle Grouping. On shape or angle 203

Female Male

Family tree Family, home town, and place of birth in relation to each other as a hierarchical network.

Home town

Anna

Marie

James

Elisabeth

Marie Jr.

Otto

Paul

Hermann

Emma

Hermann Jr.

Place of birth

Elisabeth was born as a member of the Brown family but adopted the Wagner name through her marriage. 204

Example 21

Place of birth Home town 2C.2

Family

Place of birth Home town

Gender

Relocation

3A.1

3B.1

3F.1

3F.8

Wagner Brown

London Munich Paris

Vienna

Female Male

Family Home Town 4A.1

4B.1

Sorted by age

Grouped by family and home town

Elements used 2C.2 3A.1 3B.1 3F.1 3F.8 4A.1 4B.1

Relationship. Network. Hierarchical Color. Basic colors Shape. Geometrical Isotype. Figurative shape Isotype. Background. Figurative Sorting. Linear axis Grouping. Linear 205

Yes

Chain Family set in relation as linear network and sorted linearly by year of birth. Age group as area.

Hermann

Emma

Marie

James

Otto

Anna

Paul

Elisabeth

Hermann Jr.

Marie Jr.

In the Wagner family, the birth of a female member was always preceded by the birth of a male member. 206

Age group

Family

2A.5

2C.3

Example 22

S  85

Family

Gender

Relocation

3A.1

3B.6

3F.5

Wagner Brown

Female Male

Yes No

4A.1

Sorted by year of birth

Elements used 2A.5 2C.3 3A.1 3B.6 3F.5 4A.1

Quantity. Area size Relationship. Network. Linear Color. Basic colors Shape. Blurring Isotype. Background / Foreground Sorting. Linear axis 207

Skyline Relocation and home town set in relation to age as a circular network. Age as line length.

a Ann (83)

Emma (85)

Herm a (90) nn

Pa (97 ul ) El h et ab ) is (99

Marie Jr. (78)

Marie (72)

James (65)

Hermann Jr. (52)

Otto (23)

Only Wagners lived in Vienna and Munich. 208

Age

Home town

2A.1

2C.4

Example 23

6.25 mm = 10 years

Family

Home town

Gender

Relocation

3A.1

3B.1

3C.2

3C.5

Wagner Brown

London Munich Paris

Female Male

Vienna

4A.6

Sorted by age

Elements used 2A.1 Quantity. Line length 2C.4 Relationship. Network. Circular 3A.1 Color. Basic colors 3B.1 Shape. Geometrical 3C.2 Line. Thickness 3C.5 Line. Details 4A.6 Sorting. Radially. Evenly distributed on form 209

Yes No

Intersection Place of birth and home town set in relation to each other as a nested section. Age as area.

Paris

Vienna

Elisabeth (99)

Emma (85)

Hermann (90)

Marie (72) Hermann Jr. (52) Paul (97)

James (65) Marie Jr. (78)

Otto (23)

Anna (83) Munich

London

Two Wagners relocated from Paris, but one Brown relocated to Paris. 210

Example 24

Age group

Place of birth Home town

2A.5

2C.5

S  85

Overlap = relocation

Family

Relocation

Generation

Gender

3A.1

3A.6

3B.4

3D.3

Wagner Brown

Overlap = relocation

1st 2nd 3rd

4A.9

Sorted free in space

Elements used 2A.5 2C.5 3A.1 3A.6 3B.4 3D.3 4A.9

Quantity. Area size Relationship. Nesting. Intersection sets Color. Basic colors Color. Transparency Shape. Direction Pattern. Line thickness Sorting. Free without axis or point 211

Female Male

Cells Place of birth and home town set in relation to each other as nested stacks and grouped on levels by place of birth.

Hermann Emma

Paul

Otto

Vienna

Marie

Elisabeth

Hermann Jr. Marie Jr.

Paris

Anna James

London

Half of all people living in Paris and London relocated, whereas only a quarter of those living in Vienna did so. 212

Example 25

Place of birth Home town 2C.6

Outside: Place of birth Inside: Home town

Family

Gender

Age group

Place of birth Home town

3A.1

3A.2

3B.5

3E.5

Wagner Brown

Female Male

2:  85

Place of birth 4A.9

4B.6

Sorted free in space

Grouped by place of birth

Elements used 2C.6 3A.1 3A.2 3B.5 3E.5 4A.9 4B.6

Relationship. Nesting. Stacking Color. Basic colors Color. Saturation with white Shape. Stacking Contour. Details Sorting. Free without axis or point Grouping. On levels 213

London Munich Paris

Vienna

Cluster Generation and home town set in relation to each other as a nested matrix and sorted by generation and home town.

Munich

Paris

London

Vienna

First generation James

Marie

Hermann Emma

Second generation Otto

Anna

Elisabeth

Paul

Third generation Marie Jr.

Hermann Jr.

Mobility increases with each generation. 214

Example 26

Home town Generation 2C.7

Family

Relocation

Gender

3A.1

3F.3

3F.4

Wagner Brown

Left = Yes Right = No

Female Male

Elements used 2C.7 3A.1 3F.3 3F.4

Relationship. Nesting. Matrix Color. Basic colors Isotype. Direction Isotype. Background. Shaded 215

6 Appendix

217

The authors Lucerne School of Art and Design

Pioneer of DX — D ata Experience Design

Over 100 projects for customers such as Roche, Schweizer Post, nongovernmental organizations such as UNICEF, the UN, and government institutions such as the Swiss Statistics Office and the Swiss National Science Foundation.

Initiator of the event series “On Data And Design” Basel —  Berlin — Vienna

Superdot.studio Information is Beautiful Award DPA Infographics Award Out of Balance Kritik der Gegenwart Information design winner

YAAY.ch

Training

Career

Highlight 218

FHNW Academy of Art and Design Basel Bern Academy of the Arts University of Applied Sciences of the Grisons Teaching and mentoring

Darjan Hil

Nicole Lachenmeier FHNW Academy of Art and Design Basel

University of Vienna BSc Business Informatics

BA Visual Communication (i.a. with Wolfgang Weingart)

MSc Business Informatics

The most beautiful Swiss books

Oracle / IBM Certified Systemic Coach Credit Suisse

Research at FHNW Academy of Art and Design Basel Studio Niclac

MA Visual Communication and Iconic Research Basel

Research at FHNW Academy of Art and Design Basel Research at FHNW School of Business Basel 219

Further reading, sources, and tools Books that have accompanied us along the way and that we recommend as further complementary reading to this book.

Swiss graphic design

Information Design Journal

Maier, M. (1986). ‘Elementare Entwurfs- und Gestaltungs­ prozesse: Die Grundkurse an der Kunstgewerbeschule Basel’. Bern.

R. Klanten, N. Bourquin, et al. (2008). ‘Data Flow: Visualizing Information in Graphic Design’ (1st Ed.). Berlin.

Müller-Brockmann, J. (2009). ‘Grid systems in Graphic Design: A Visual Commu­ nication Manual for Graphic Designers, Typographers and Three Dimensional Designers’ (4th Rev. Ed.). Salenstein. Osterer, H. & Stamm, P. (2008). ‘Adrian Frutiger — Typefaces: The Complete Works’ (1st Ed.). Basel.

Visual perception Dondis, D. A. (1974). ‘A Primer of Visual Literacy’ (New Ed.). Cambridge, Massachusetts. Leborg, C. (2006). ‘Visual Grammar (Design Briefs)’ (1st Ed.). New York City. Zwimpfer, M. (1994). ‘2d — visuelle Wahrnehmung’. Salenstein.

Schwochow, J. (2020). ‘Die Welt verstehen mit 264 Infografiken: aus Geschichte, Wirtschaft, Gesellschaft, Sport, Kunst und Kultur, Wissenschaft und Technologie’. Munich. Wiedemann, J. & Rendgen, S. (2018). ‘Understanding the World. The Atlas of Infographics’ (Multilingual ed.). Cologne. Wiedemann, J. & Rendgen, S. (2019). ‘History of Information Graphics: HISTORY OF INFOGRAPHICS’ (Multilingual ed.). Cologne.

Information architecture: the L.A.T.C.H theory Bradford, P. & 2 Wurman, R. S. (1996). ‘Information Architects’. New York City. 2 Wurman, R. S. (1989). ‘Information Anxiety’ (1st Ed.). Doubleday. New York City.

220

2 Wurman, R. S. (2017). ‘Understanding Understanding’. Richard Saul Wurman & Jack. New York City.

1 Otto Neurath and Jacques Bertin Bertin, J. (2011). ‘Semiology of Graphics’ (1st Ed.). Madison. Brian, S., Neurath, M. & Kinross, R. (2016). ‘Die Transformierer: Entstehung und Prinzipien von Isotype’ (1st Ed.). Salenstein. Hartmann, F. & Bauer, E. K. (2006). ‘Bildersprache: Otto Neurath Visualisierungen’ (2nd ext. Ed.). Vienna. Vossoughian, N. (2011). ‘Otto Neurath: The Language of the Global Polis’ (1st Ed.). Rotterdam.

Theory of information design Tufte, E. R. (1990). ‘Envi­sioning Information’ (1st Ed.). Cheshire, Connecticut. Tufte, E. R. (1997). ‘Visual Explanations: Images and Quantities, Evidence and Narrative’. Cheshire, ­C onnecticut.

Tufte, E. R. (2001). ‘The Visual Display of Quantitative Information’ (2nd Ed.). Cheshire, Connecticut. Tufte, E. R. (2006). ‘Beautiful Evidence’. Cheshire, Connecticut.

Information design manuals

Recommended tools for the visualization process For the analysis of data: Excel and Tableau; for the implementation: Illustrator, InDesign, RAW Graphs, Processing / p5.js, SVG with Vanilla JS, D3.js

Kirk, A. (2019). ‘Data Visualisation: A Handbook for Data Driven Design’ (Reprint Ed.). New York City. Koponen, J. & Hildén, J. (2018). ‘Data Visualization Handbook’. Helsinki. Lima, M. (2013). ‘Visual Complexity: Mapping Patterns of Information (History of Information and Data Visualization and Guide to Today’s Innovative Appli­ cations)’ (Reprint Ed.). New York City. Lupi, G. & Posavec, S. (2018). ‘Observe, Collect, Draw!: A Visual Journal’ (1st Ed.). New York City. Schwabish, J. (2021). ‘Better Data Visualizations: A Guide for Scholars, Researchers, and Wonks’. New York City.

221

Acknowledgments

No book is ever written overnight, and it is never the work of just one person. Unfortunately, there are not enough lines to mention all of our friends and family members who were directly or indirectly involved in this book, even if only through motivating words or unconscious influence. However, we want to express our particular thanks to two people from our Superdot team who have given us many ideas and much support — Sarah Heitz and Aaron Ritschard; without your trust, your patience, and your contribution, we would not have managed. Many thanks also to Birkhäuser Publishers for ­trusting in our work. # visualizingcomplexity # modularinformationdesign

222

Publication details

Concept and content

Darjan Hil, Nicole Lachenmeier

Layout and design Nicole Lachenmeier

Typesetting

Superdot • visualizing complexity

Translation from German into English

Hartwin Busch

Copyediting Patricia Kot

Project management Baharak Tajbakhsh, Freya Mohr

Production

Amelie Solbrig, Heike Strempel

Paper

Juwel Offset

Printing

Eberl & Kœsel GmbH & Co. KG Library of Congress Control Number: 2021953442

right. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, re-use of illustrations, recitation, broadcasting, reproduction on microfilms or in other ways, and storage in databases. For any kind of use, permission of the copyright owner must be obtained. ISBN 978-3-0356-2504-2 e-ISBN (PDF) 978-3-0356-2506-6 German Print-ISBN 978-3-0356-2505-9 © 2022 Birkhäuser Verlag GmbH, Basel P.O. Box 44, 4009 Basel, Switzerland Part of Walter de Gruyter GmbH, Berlin / Boston 9 8 7 6 5 4 3 2 1 www.birkhauser.com

­

Bibliographic information published by the German National Library. The German National Library lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at http://dnb.dnb.de.  This work is subject to copy223