Principles of infographics

Question 1

3 Characteristics of a good infographics

Answer 1

Utility
Soundness
Beauty

Question 2

Utility

Answer 2

With respect to utility, infographics must employ an objectives based approach. Essentially, the utility of an infographic is measured by how it enables a brand to reach its objectives.

Question 3

The utility of explorative and narrative infographics

Answer 3

explorative infographics provide information in an unbiased fashion, enabling viewers to analyse it and arrive at their own conclusions. This approach is best used for scientific and academic applications, in which comprehension of collected research or insights is a priority. Narrative infographics guide the viewers through a specific set of information that tells a predetermined story. This approach is best used when there is a need to leave readers with a specific message to take away, and should focus on audience appeal and information retention.

Question 4

Soundness

Answer 4

Good infographics also communicate something meaningful. Communicating a message worth telling provides readers with something of value. While infographics can be a powerful vehicle of communication, they are sometimes produced arbitrarily or when a cohesive and interesting story isn’t present. If the information itself is incomplete, untrustworthy, or uninteresting, attempting to create a good infographic with it is more than a fool’s errand; it’s impossible.

Question 5

Beauty

Answer 5

how it is designed—is also important. With this in mind, there are two things to consider: format and design quality. If an inappropriate format is used, the outcome will be inferior. Similarly, if the design misrepresents or skews the information deliberately or due to user error, or if the design is inappropriate given the subject matter, it cannot be considered high quality, no matter how aesthetically appealing it appears at first glance.

Question 6

Illustration

Answer 6

“[Illustration] can help showcase and bring things to life, but there’s a fine line between [this and] it becoming distracting. It should be more of a supporting than a leading role. If the designer isn’t careful, what he’s designing can become purely an illustration, rather than an infographic. You want the illustration to support the story that the data is telling, rather than detract from it

Question 7

Illustration appropriateness in narrative and explorative approaches

Answer 7

For example, illustration can distract and would not be appropriate for an explorative infographic, because people will spend time looking at the visuals while trying to determine what the message is. As we’ve mentioned, the sequence should always be information first, design second. However, the message in a narrative infographic has already been established. Therefore, this is a case where good illustration is capable of playing that supporting role we referenced earlier.

Question 8

When else can illustration be harmful

Answer 8

Illustration can also be harmful if it misleads people when used improperly, not because illustration is innately misleading. The most common mistake we encounter is the accidental distortion of the data’s display. We also frequently see people who use illustration to hide the fact that their message is either incomplete or meaningless.

Question 9

4 main types of relationships displayed by infographics

Answer 9

Nominal comparison
Time series
Ranking
Part to whole

Question 10

Nominal comparison

Answer 10

Nominal comparisons represent a nominal scale; their function is to display several subcategories’ quantitative values so that they can be easily compared to each other. An example of the type of message that ought to be conveyed through this type of comparison is that the value of X is larger than that of Y, or that the value of B is twice as large as the value of C.

Question 11

Best way to show a nominal comparison

Answer 11

Quantitative values in a nominal comparison graph are represented by categorical subdivisions (eg. B or C) are independent of one another. They should therefore be designed in such a way that communicates the uniqueness of each value. Bar graphs work really well for this type of relationship—especially if the scale is large and you want to highlight the differences between the values. You can also shorten the range and replace bars with simple data points, known as a dot plot.

Question 12

Time series

Answer 12

A time series relationship consists of a number of relationships between categorical quantitative values that are distributed across divisions of time. Time series graphs are used to show trends, or how values change over time. They take a single measure for each line and plot change (positive and negative) in its quantitative value over time. These are the most popular types of graphs used in business

Question 13

Best way to show time series

Answer 13

The most commonly used graphs to show time series relationships are lines (with continuous data) and vertical bars (with discrete data). Dot plots can also be used, as well as a dot plot with lines connecting the various points. You’ll want to avoid horizontal bar charts. We typically represent time as something that moves from left to right, rather than top-down.

Question 14

Ranking

Answer 14

Graphic ranking relationships are all about communicating the ordering—from highest to lowest, or vice versa—of quantitative values of a set of subcategories

Question 15

Best way to show ranking

Answer 15

If the goal is to show ranking, then you want to use a graph type that brings attention to the quantitative values of each subcategory. Bar graphs (vertical and horizontal) are extremely useful for visualizing ranking relationships; the ordering of categorical subdivisions should be selected according to the purpose of highlighting the highest values (descending order) or the lowest values (ascending order). Dot plots can also be used, if narrowing the quantitative scale aids in the display of the ranking relationship.

Question 16

Part to whole

Answer 16

The goal in graphing a parts-to-whole relationship is to show how a set of categorical subdivisions’ quantitative values relate to one another as parts of a whole. By displaying a ratio, a part to-whole graph uses percentages as the unit of measure. When graphing part-to-whole relationships, the sum of all individual quantitative values must always add up to 100 percent. A common example of when part-to-whole relationships can be used is when people want to show budget breakdowns; essentially what percentage of a budget is used for what.

Question 17

How to best show part to whole

Answer 17

Pie charts are the most common graph types used for partto-whole relationships; however, they have their limitations. You want to avoid using a pie chart if you have more than a handful of categorical subdivisions. A stacked bar chart is also a good option; this is especially true if you would like readers to be able to compare the composition of two separate stacked bars, with the same categorical subdivisions. For example, if you wanted to show the favourite colours of students from three separate classrooms. Bar (horizontal and vertical) graphs and dot plots can be useful as well, but both are less popular than those previously mentioned for this type of relationship.

Question 18

6 main types of graphs

Answer 18

• Dot plot
• Line chart
• Vertical bar graph
• Horizontal bar graph
• Stacked bar graph
• Pie chart

Question 19

Important rule in regards to dot plots

Answer 19

When using dot plots to show a tie series relationship the scale does not have to start at a 0 baseline but for other relationships they do.

Question 20

Dot plots

Answer 20

Dot plots can be used with either discrete or continuous data. They essentially consist of a set of dots plotted along an x-axis according to qualitative values (e.g., subcategories), and vertically on the y-axis according to quantitative value.

Question 21

Dot plots and time series

Answer 21

Dot plots are most commonly used to show time series relationships, with the various points along the x-axis representing sequential points in time. In this case, the y-axis could be thought of as the “value axis” and the x-axis could be referred to as the “time axis.” Trends can be identified based on a change in height of dot placement along the x-axis.

Question 22

Dot plots examples

Answer 22

Photo 1

Question 23

Line charts

Answer 23

We use line charts to show time series relationships with continuous data (Figure 9.6). As could probably be imagined, your line chart could have the dots at specified intervals (days, months, years, etc.) or it can omit them. With line graphs, the x-axis should always represent time, while the y-axis should represent a quantitative value that changes over time.

Question 24

Line charts are very handy

Answer 24

they allow users to identify specific values at various points in time, as well as to identify trends, such as when and how much a value is changing. Charting multiple values on the same graph allows readers to further identify the relationship between categories; for instance, an increase in one variable and a decrease in another, or even an increase or decrease across all variables.

Question 25

Dot plots and line charts scaling

Answer 25

As with dot plots, the scale on line charts has a lot to do with how the message is conveyed. For example, using too large a scale runs the risk that viewers may gloss over a very important story in the data. However, using too small a scale might lead you to overemphasize minor fluctuations. As with dot plots, designers should plot all of the data points so that the line chart takes up two-thirds of the y-axis’s total scale.

Question 26

Panelling

Answer 26

Most people would argue that line charts are very easy to understand, and should be kept that way. Showing too many lines on one chart can tend to look too busy; therefore, it is best to keep the chart to four or fewer clearly labelled lines. If you need to show more than four categories, you can use the practice of panelling (Figure 9.7) and a constant scale for consistency.

Question 27

Line charts examples

Answer 27

Photo 2

Question 28

Bar graph

Answer 28

Bar graphs can use discrete or continuous data, and there’s really not a lot that you can’t do with them. Similar to line charts, vertical bars have qualitative values listed on the x-axis and quantitative values on the y-axis. They can also be rotated 90 degrees for horizontal placement (more below), upon which the axes’ value types would be swapped.

Question 29

The value of a bar graph

Answer 29

The value of a bar graph is that it makes it easy for the reader to understand what is going on. You should resist all urges to overcomplicate things. The first rule about bar graphs is that they must always have a zero baseline. This is because we use the length of the bars to compare them against each other, and if we truncate the scale, the story becomes distorted

Question 30

Other bar graph rules

Answer 30

According to Dona Wong, the perfect distance between two bars in a graph is half the width of one of the bars (Wong, p. 63). Typically each bar within a graph should be the same color, and you should avoid any intense patterns if it distracts from understanding the visualization.

Question 31

Bar graph examples

Answer 31

Photo 3

Question 32

Limitation of horizontal bar graphs

Answer 32

One limitation of horizontal bars, however, is that they are not good for showing time series relationships. While they could theoretically be used for this, it’s not recommended because we are accustomed to reading about time from left to right (at least in the West).

Question 33

Major rule for horizontal bar graphs

Answer 33

to use leftward horizontal bars when the values are actually positive. The x-axis starts at zero in horizontal bar charts; thus any bar to the left of that baseline might be viewed as having a negative value. This mistake is very common among web-based infographics, and it tells the opposite story than what is actually found in the data.

Question 34

Horizontal examples

Answer 34

Photo 4

Question 35

Stacked bar

Answer 35

Stacked bars are most often used when there is a need to display multiple part-to-whole relationships. Stacked bars use discrete or continuous data, and can be oriented either vertically or horizontally. While the aggregate of each bar can be used to make nominal or ranking comparisons, this graph type is used when the composition of each bar tells an interesting story that provides the viewer with greater insight

Question 36

100 percent stacked bar

Answer 36

all of the subcategories add up to 100 percent (as is the case with the pie chart), therefore removing this nominal comparison between bars. Instead, this allows the viewer to focus on the comparative composition of each bar without the distraction of relative scale. With 100 percent stacked bars, the size of each segment represents a percentage of the total bar and is typically labelled accordingly

Question 37

100 percent stacked bar use

Answer 37

One hundred percent stacked bars are typically used when there is a need to show a time series relationship as well, (e.g., how one’s donut consumption compares to that of his friends) (Figure 9.16). Most would argue that this type of visualization is easier to understand than multiple pie charts as it enables the viewer to directly compare segments on the same axis with relative ease.

Question 38

Stacked bar examples

Answer 38

Photo 5

Question 39

The real value of pie charts

Answer 39

Making part to whole comparisons. Useful in communicating big ideas quickly.

Question 40

Limits of pie charts =

Answer 40

they’re not very useful in comparing the values of the subcategories between pies (as stacked bars can be), or showing the changing makeup of a part-to-whole relationship over time. This is because it’s hard to compare the sizes of multiple pie “slices” (essentially the angles of their points next to each other) in the same pie or across multiple pies.

Question 41

Pie chart rules

Answer 41

There are several rules to consider when using pie charts, the most important of which is that the sum of all the subcategories’ quantitative values must always equal 100 percent—no exceptions. And as we’ve previously stated, you should really only use pie charts when you have a handful of subcategories

Question 42

Pie chart arrangement =

Answer 42

When arranging the parts of the pie chart, the largest section should always start at the top, and go clockwise from 12 o’clock. Similarly, the second largest section should always start at the top and go counter clockwise from 12 o’clock. Any additional sections should be placed down below. The logic behind this is that readers read top down, and will therefore read the most important subcategories (arguably the largest) first

Question 43

Pie chart example

Answer 43

Photo 6