ized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized

TRAINING
GENDER
STATISTICS

DATA VISUALIZATION
STRENGTHENING
Contents
Data visualization tip sheet
Module 1. Introduction to data visualization
Module 2. Data visualization principles and concepts
Module 3. Excel data and charts adjustments
Module 4. Visualization tools comparison
Module 5. Excel chart transformations
Module 6. DATAWRAPPER maps
Module 7. DATAWRAPPER range plots
Module 8. Annotations
 WORLD BANK SGS DATA VISUALIZATION TRAINING | DATA VISUALIZATION TIP SHEET

Chart type When to use Chart-specific tips

Column chart Bar chart • Mostly for one variable • Should use the same color for all bars (unless
(vertical bar) (horizontal bar) • Show comparison/ranking specifically disaggregated for example by sex).
• Overall, Total, or National data points can be
• Bar used when variable/category labels are too long to highlighted by a different color or transformed from
show horizontally in a column chart a bar to a highlight line.
• The width of the bars should be about twice the
width of the space between the bars.

Clustered/grouped bar or column • Mostly for breaking/grouping one variable into different • Should use the same color for all bars (unless
chart levels of disaggregation (subgroups) specifically disaggregated for example by sex).
• Show comparison/ranking • Make sure multiple levels of disaggregation or group
are shown in the axis to avoid labeling each category
• Bar used when variable/category labels are too long to name individually.
show horizontally in a column chart • The gap between the clusters/groups of bars should
be half the size of the width of the bar itself.

Stacked bar or column chart • Show the composition usually out of 100% (part- to-whole • Should have no more than 8 categories.
relationship of a variable’s categories) • Legends should be stretched across the top of the
• Can be simple (one variable) or chart or to the right, and the order should match the
clustered/grouped (different levels of order in the chart.
disaggregation or subgroups) • Should use categorical color palette.
• The “Other” category should not be the largest
section of the stacked chart. Try breaking the
“other” section into smaller sections.
 WORLD BANK SGS DATA VISUALIZATION TRAINING | DATA VISUALIZATION TIP SHEET

Line chart • Show the trend in variables usually over time • No more than 4 to 6 lines in a chart.
(but can also be age ranges, or hours in a day). • When too many lines, break into “small
• Show multiple variables with multiple lines (if multiples” or “grid of charts” highlighting the lines
they are on the same scale). individually or highlighting the main line in one color
• Show the same variable for multiple observations and all other lines behind in light grey.
with multiple lines. • Option to shade area between male and female line
to better show gaps.
• When possible, directly label series. If lines are too
close together, use a legend.
• Avoid individual data labels. Individual data labels
may be acceptable if there are few data points and
data labels aren’t overlapping.

Area chart • Show the trend in composition (part-to-whole • Should use categorical color palette.
relationship of categories) over time. It is a combination • The “Other” category should not be the largest
of the line chart and stacked column chart. section of the stacked chart. Try breaking the “other”
section into smaller sections.
• X-axis should have an ordinal variable (time, age,
hours, temperature, etc.)

Pie chart or donut chart • Show the composition out of 100% (part-to- whole • Avoid pie charts whenever possible and
relationship of categories) typically for 3 or 4 but preferably use stacked charts instead.
no more than 6 categories. • If it is necessary to use a pie chart, there should be
no more than 6 slices.
• The “Other” slice should not be the largest
slice of the pie chart.

Range plot • Show comparison/ranking • Order the range plot by value for ease of
• Show gaps and absolute values comprehension.
• Gridlines should always be removed to avoid the
length of the gap blending into the gridline.
 WORLD BANK SGS DATA VISUALIZATION TRAINING | DATA VISUALIZATION TIP SHEET

Heatmap • Show comparison with multiple levels or • For continuous data, use sequential color palette or
disaggregation with many subgroups or categories if showing the gender gap, use a diverging color
• Most commonly shown over time (hours, years) or across palette.
age groups • X-axis should have an ordinal variable (time, age,
• Pack a lot of information in one chart hours, temperature, etc.)
Scatterplot • Show relationship between two variables • Is typically reserved for visualizing data with many
• Could show the relationship between male and female observations (i.e., microdata).
values of the same variable or between two years of same • Add a 45-degree line to show equal values between
variable. the two variables. For example, gender parity when
• Show distribution, outliers plotting male and female values on the axes.
• Mostly for microdata but could plot disaggregation • Using a third variable as dot (bubble) size creates a
category with many subgroups (i.e., geographic regions, bubble chart.
occupations)
Dot plot • Show distribution of the whole set of • Is typically reserved for visualizing data with many
observations observations (i.e., microdata).
• Easily identify outliers • Can be shown vertically or horizontally.
• Mostly for microdata but could plot disaggregation • Can be colored by one of the disaggregation
category with many subgroups (i.e., geographic regions, categories. For example, plotting all households
occupations) from microdata and coloring by urban/rural.
Map • Show geographic data much more visually than through the • Should use sequential colors unless showing a gap
aforementioned chart types. which requires a diverging color scale.
• Can be shown at whichever level data are available, • Must use the same color scale and range/steps for
national, administrative regions, provinces, districts, male and female for proper side by side comparison.
cities, etc. • For a gap, use text in the legend instead of numbers
so as not to confuse the reader with negative
numbers. Legend should show parity in the middle
and female values higher than male values at one
end and vice versa for the opposite end of the scale.
 WORLD BANK SGS DATA VISUALIZATION TRAINING | DATA VISUALIZATION TIP SHEET

DATA VISUALIZATION CHECKLIST

Overall Does the visual highlight a significant gender-relevant finding or conclusion? If not, can the data structure be adjusted to highlight the
gender-relevant insight or is the data not gender-relevant?
Graphs will catch a
Total or aggregate statistics can often be removed from tables and graphs to facilitate comparisons between women and men. If the totals
viewer’s attention so only
are large, choose a chart type that highlights overall and gender-disaggregated statistics.
visualize the data that needs
attention. Too many graphics of
unimportant information dilute
the power of visualization. Is the type of visual appropriate for the data?
For example, change over time is displayed as a line graph, area chart, slope graph, or dot plot.

Does the visual have appropriate level of precision?

Use a level of precision that meets your audiences’ needs. Few numeric labels need decimal places, unless you are speaking
with academic peers. Charts intended for public consumption rarely need p values listed.

Do the individual visual elements work together to reinforce the overarching takeaway message? Are any of the visual elements
duplicating information?
Choices about graph type, text, arrangement, color, and lines should reinforce the same takeaway message without
duplication.

Color Is the color scheme intentional? Is there a consistent color for male and female?
Colors should be derived from an intentional choice, not the default color schemes. Use your organization’s colors.
Keep culture-laden color
Avoid colors and figures that reinforce gender stereotypes such as pink for women/girls and blue for men/boys.
connotations in mind.
Always use one color to represent women and another color to represent men. These colors should not be two different shades of the same color hue
Use sites like Color (i.e., light blue and dark blue). They should be two distinct shades of color (i.e., blue and orange).
Brewer to find color
schemes suitable for
Does the color highlight key patterns?
reprinting in black- and-
Action colors should guide the viewer to key parts of the display. Less important, supporting, or comparison data should be
white and for
a muted color, like gray.
colorblindness.
Is the color legible when printed in black and white or for people with colorblindness?
When printed or photocopied in black and white, the viewer should still be able to see patterns in the data. Avoid red-green and yellow-blue combinations when
those colors touch one another. Avoid using red to mean bad and green to mean good in the same chart.
 WORLD BANK SGS DATA VISUALIZATION TRAINING | DATA VISUALIZATION TIP SHEET

DATA VISUALIZATION CHECKLIST

Color (continued) For accessibility, don't rely solely on color to convey the idea. As much as possible, the presentation of the data should still be understandable if the color is
removed or there should be enough additional text that identifies the key ideas in the table for the visually impaired audience who can read the report with
a screen reader.

Does the text sufficiently contrast the background?

Black/very dark text against a white/transparent background is easiest to read.

Lines Are there gridlines? If yes, are they muted?

Color should be faint gray, not black. There should be a preference for no gridlines. Gridlines, even muted, should not be
Excessive lines— gridlines, used when the graph includes numeric labels on each data point.
borders, tick marks, and Does the visual have a border line?
axes—can add clutter or noise Graph should bleed into the surrounding page or slide rather than being contained by a border.
to a graph, so eliminate them
whenever they aren’t useful Do the axes have unnecessary tick marks or axis lines?
for interpreting the data. Tick marks can be useful in line graphs (to demarcate each point in time along the y-axis) but are unnecessary in most
other graph types. Remove axes lines whenever possible.

Does the graph have one horizontal and one vertical axis?
Viewers can best interpret one x-axis and one y-axis. Don’t add a second y-axis. Try a connected scatter plot or two graphs, side by side, instead. (A
secondary axis used to hack new graph types is ok, so long as viewers aren’t being asked to interpret a second y-axis.)

Arrangement Are the proportions accurate?

A viewer should be able measure the length or area of the graph with a ruler and find that it matches the relationship in
Improper arrangement of the underlying data. Y-axis scales should be appropriate. Bar charts start axes at 0. Other graphs can have a minimum and maximum scale that
graph elements can confuse reflects what should be an accurate interpretation of the data.
readers at best and mislead
viewer at worst. Thoughtful
arrangement makes a data Are the data intentionally ordered or sorted?
visualization easier for a viewer
to interpret.
 WORLD BANK SGS DATA VISUALIZATION TRAINING | DATA VISUALIZATION TIP SHEET

DATA VISUALIZATION CHECKLIST

Arrangement Data should be displayed in an order that makes logical sense to the viewer. Data may be ordered by frequency counts (e.g., from greatest to
least for nominal categories), by groupings or bins (e.g., histograms), by time period (e.g., line
(continued) charts), alphabetically, etc. Use an order that supports interpretation of the data.

As a general rule, the chart should be ordered by value (for sex-disaggregated data typically by female value). Female values should always appear
first in a graph compared to male values (above or to the left of the values for men) – this is the case for bars or columns, for the order of tables, and
also for legends.

Are axis intervals equidistant? If not, is there a symbol that signifies the jump?
The spaces between axis intervals should be the same unit, even if every axis interval isn’t labeled.
Irregular data collection periods can be noted with markers on a line graph, for example. Some charts should not start at 0 or we will not see a
significant pattern, but there should at least be a symbol to make it clear that it does not start at 0).

Is the graph two-dimensional?

Avoid three-dimensional displays, bevels, and other distortions.

Text Do subtitles and/or annotations provide additional information?

Graphs don't contain much

Subtitles and annotations (call-out text within the graph) can add explanatory and interpretive power to a graph. Use them
text, so existing text must
to answer questions a viewer might have or to highlight specific data points.
encapsulate your message
and pack a punch. Is the text size hierarchical and readable?
 WORLD BANK SGS DATA VISUALIZATION TRAINING | DATA VISUALIZATION TIP SHEET

DATA VISUALIZATION CHECKLIST

Titles are in a larger size than subtitles or annotations, which are larger than labels, which are larger than axis labels, which are larger than source information. The smallest text -
Text axis labels - are at least 9 point font size on paper, at least 20 on
(continued) screen.

Is the text horizontal?

Titles, subtitles, annotations, and data labels are horizontal (not vertical or diagonal). Line labels and axis labels can deviate
from this rule and still receive full points. Consider switching graph orientation (e.g., from column to bar chart) to make text horizontal. For cultures that write vertically, make
sure to follow cultural language practices instead.

Are data labeled directly?

Whenever possible, position data pr category/series labels near the data rather than in a separate legend (e.g., on top of or
next to bars and next to lines). Eliminate/embed legends when possible because eye movement back and forth between the legend and the data can interrupt the brain’s
attempts to interpret the graph.

Are labels used sparingly?

Focus attention by removing the redundancy. For example, in line charts, label every other year on an axis. Do not add
numeric labels *and* use a y-axis scale, since this is redundant. Don't clutter the labels, make sure there is sufficient and consistent space between the labels.

Are the numbers adjusted correctly?

Numbers rounded to no more than one decimal place, unless it is absolutely necessary to have more decimal places. When viewing large numbers, display the data in thousands
or millions and specify the unit in the subtitle or axis label. The abbreviation can also be used (in English: K for thousands, M for millions, B for billions). Don't compare fractions
with different denominators, use percentages instead.

Does the direction of the text respect cultural and linguistic rules?
Make sure that graphs and tables are aligned according to the rules for right-to-left or top-down languages. Make sure that the correct grammar, punctuation, and
capitalization follow the rules of the corresponding language. When using English, make sure that the graphics and text are consistent with only one type of English (American
or British English).

Adapted from "Data Visualization Checklist" by Stephanie Evergreen & Ann K. Emery
 WORLD BANK SGS DATA VISUALIZATION TRAINING | DATA VISUALIZATION TIP SHEET
Border

Anatomy of a visual 3D

Label
Chart Title
89
REGION G 44
67
Data Label
56
REGION F 42
49
51
REGION E 48
50
48
REGION D 41
44
49
REGION C 46
48
45
REGION B 42
43
51
REGION A 53
52
Axis
0 10 20 30 40 50 60 70 80 90 Gridlines
Internet Access

Internet Access Female Internet Access Male Internet Access Total

Axis Title
Axis Label
Legend
STRENGTHENING
GENDER
STATISTICS

DATA VISUALIZATION
TRAINING MODULE 1:
INTRODUCTION TO
DATA VISUALIZATION
Contents
1. What is data visualization?
2. Why is data visualization important?
3. Advantages/disadvantages of data visualization.
4. How are data visualizations disseminated?
5. For whom are data visualizations produced?
6. Practice exercise
1. Introduction to
data visualization
 What is data visualization?
➢ Data visualization is the graphical representation of information and data.

➢ Visual elements:
• Include charts, diagrams, data visualization attributes (shape, position, color, pattern, etc.),
images/icons, annotations (lines, shaded areas, arrows, etc.)
• Provide an accessible way to see and understand trends, outliers, and patterns in data.

➢ Data visualization involves human perception & cognition.

• The human mind is slow with mental operations like multiplying, subtracting, or simply comparing
numbers.
• Data visualizations present information that we can visually perceive to better understand the
insights and trends in the data.

“Data analytics is the data representation and presentation that

exploits our visual perceptions capabilities to amplify the cognition.”

- Andy Kirk, author of “Data Visualization: a successful design process”

 GROUP DISCUSSION
➢ What insights does the table provide?
➢ Can you easily tell which level of education has the highest enrollment?
➢ Can you easily compare enrollment for different cities? From year 1 to year 2?

Pre-primary Pre-primary
education Primary education Secondary education enrollment Y1 to
Y2 Improvement
Y1 Y2 Y1 Y2 Y1 Y2
City A 11.1 18 66.4 77 19.9 32.2 6.9
City B 7.5 15.2 62.8 73.8 14.4 23.1 7.7
City C 23.6 26.6 77.7 86.8 35.2 59.2 3
City D 10.9 17.5 66.2 76.8 19.7 31.4 6.6
City E 17.7 25.8 81.9 92.7 42.1 64.2 8.1
City F 28.2 26.4 77.1 84.5 33.2 55.4 -1.8
City G 7.2 14.7 62.6 73.5 13.9 22.8 7.5
City H 18.5 36.1 71.8 85 27.5 61 17.6
City I 16.7 32.5 64.6 76.5 24.8 54.9 15.84
City J 9.7 17.9 62.5 75.6 18.3 29.8 8.2
City K 12.8 18 70.6 78.5 21.6 34.6 5.2
 GROUP DISCUSSION
➢ What insights does the chart provide?
➢ Can you easily tell which level of education has the highest enrollment?
➢ Can you easily compare enrollment for different cities? From year 1 to year 2?
➢ Is it easier to answer these questions compared to the table?
Simplified presentation
1) Too much information is 2) Non intuitive for the
being shared in one chart audience to read through

Pre-primary
education Primary education Secondary education Improvement

Y1 Y2 Y1 Y2 Y1 Y2
City A 11.1 18 66.4 77 19.9 32.2 6.9
City B 7.5 15.2 62.8 73.8 14.4 23.1 7.7
City C 23.6 26.6 77.7 86.8 35.2 59.2 3
City D 10.9 17.5 66.2 76.8 19.7 31.4 6.6
City E 17.7 25.8 81.9 92.7 42.1 64.2 8.1
City F 28.2 26.4 77.1 84.5 33.2 55.4 -1.8
City G 7.2 14.7 62.6 73.5 13.9 22.8 7.5
City H 18.5 36.1 71.8 85 27.5 61 17.6
City I 16.7 32.5 64.6 76.5 24.8 54.9 15.84
City J 9.7 17.9 62.5 75.6 18.3 29.8 8.2
City K 12.8 18 70.6 78.5 21.6 34.6 5.2

3) Not immediately visible

what the data is saying –
where is the main insight?
 Simplified presentation
➢ Main goal: highlight trends in pre-primary enrollment over time.
• The left chart can be simplified to show only the pre-primary enrollment → as in the right chart.
 Simplified presentation
➢ Main goal: highlight trends in pre-primary enrollment over time.
• The left chart can be simplified to show only the pre-primary enrollment → as in the right chart.

1) The most relevant information

is being shared

2) The chart is more readable but

still has too many options…
 GROUP DISCUSSION
➢ Key insight: top 3 cities where pre-primary enrollment has improved the most over time.
➢ Can you easily tell which improvement in enrollment has occurred in each city?
➢ Which are the top 3 most improved cities?
 Simplified presentation
➢ Key insight: top 3 cities where pre-primary enrollment has improved the most over time.
• Option 1: Highlight the improvement for each city sorted by improvement level.
• Option 2: Highlight only the 3 top cities but with absolute numbers for year 1 and 2.

Improvement in Pre Primary enrollment from Y1 to Y2

20 17.6
15.84
15

10 8.2 8.1 7.7 7.5 6.9 6.6

5.2
5 3

-5 -1.8
City H City I City J City E City B City G City A City D City K City C City F

Provides a better sense of how pre- Conveys the exact information we are
primary enrollment has changed. trying to communicate with no extra
Top 3 easily identifiable. details. No need to search top cities.
 Why is data visualization important?
The importance is simple – data visualization helps people:

See Interact with Better understand data

➢ Whether simple or complex, the right visualization can bring everyone on the same page, regardless
of their level of expertise.

➢ While traditional education typically draws a distinct line between creative storytelling and technical
analysis, the modern professional world also values those who can cross between the two:
Data visualization sits right in the middle of analysis and visual storytelling.

Goals of a data visualization: to explain, to monitor, and to relay information

Advantages/disadvantages of data visualization

ADVANTAGES DISADVANTAGES

• Easily identify trends that could be • Cannot always discern exact values
missed in tables. from the visual or they are not
provided.
• Appropriate use of visual elements
(colors/patterns) makes it easier to • Core messages can get lost.
find patterns and relationships.
• Wrong design or visual practices can
• Possibility to explore data in lead to inaccurate representations or
interactive formats through biased information.
dashboards.
 Advantages/disadvantages of data visualization
➢ Can you quickly identify which city has the largest population?

Population Disease
City A 47.1 28.2
City B 102.0 61.2
City C 63.8 38.3
City D 96.0 57.6
City E 98.0 58.8
City F 50.2 30.1
City G 3.0 1.8

One has to look through all

the values in the table to
decipher which one has the
largest population.
 Advantages/disadvantages of data visualization
➢ Can you quickly identify which city has the largest population?

Population Disease
City A 47.1 28.2
City B 102.0 61.2
City C 63.8 38.3
City D 96.0 57.6
City E 98.0 58.8
City F 50.2 30.1
City G 3.0 1.8

The chart makes it easier to identify

that City B and City E have the highest
populations, but the difference is very
minimal in the visual and ambiguous.
 Advantages/disadvantages of data visualization
➢ Can you quickly identify which city has the largest population?

Population Disease
City A 47.1 28.2
City B 102.0 61.2
City C 63.8 38.3
City D 96.0 57.6
City E 98.0 58.8
City F 50.2 30.1
City G 3.0 1.8

Using appropriate visual principles

like color differentiation and sorting
by value provides the information
even more quickly.
 Advantages/disadvantages of data visualization
➢ Can you quickly identify which city has the largest population?

Population Disease
City A 47.1 28.2
City B 102.0 61.2
City C 63.8 38.3
City D 96.0 57.6
City E 98.0 58.8
City F 50.2 30.1
City G 3.0 1.8

The table easily

pinpoints the largest
population by exact
data point, while the
chart without data
labels does not specify.
 Advantages/disadvantages of data visualization
➢ Can you quickly identify which city has the largest population?

Population Disease
City A 47.1 28.2
City B 102.0 61.2
City C 63.8 38.3
City D 96.0 57.6
City E 98.0 58.8
City F 50.2 30.1
City G 3.0 1.8

No Disease Disease

In this chart type, you can understand the

proportion of population with and without
disease, but you cannot tell the population of
the cities, which vary as seen in the table. The
proportions are also very similar so there isn’t
any comparable insight by city.
 Advantages/disadvantages of data visualization
➢ Can you quickly identify which city has the largest population?

Population Disease
City A 47.1 28.2
City B 102.0 61.2
City C 63.8 38.3
City D 96.0 57.6
City E 98.0 58.8
City F 50.2 30.1
City G 3.0 1.8

No Disease Disease

This chart is incorrectly stacking

incidence of disease on population data,
adding the total incidence of disease
with total population, which doesn’t
provide any useful information.
 How are data visualizations disseminated?

Report (i.e. Gender Factbook)

• Official statistical output with mostly text and analysis supported by visuals

Complementary executive summary/visual brief

• 1-2 pages long summarizing key visuals and insights from report

Infographic
• An eye-catching collection of images, icons, charts with minimal text

Written blog or data story

• An article that tells a narrative/story around the data using visuals for
general audiences to better comprehend.
• Interactive data stories include innovative visualizations with narrative.

Social media card or factoid card

• A quick snapshot (1-2 charts) of a key insight which could be taken from a
report or created with data completely separate from any other product.
Interactive visual for social media
• An interactive visual, collection of animated visuals, or data story that
allows users to interact with the chart, hover over parts of the visual to get
exact data values and information and explore the various elements
 How are data visualizations disseminated?

Report (i.e. Gender Factbook)

• Official statistical output with mostly text and analysis supported by visuals

Complementary executive summary/visual brief

• 1-2 pages long summarizing key visuals and insights from report

Infographic
• An eye-catching collection of images, icons, charts with minimal text

Written blog or data story

• An article that tells a narrative/story around the data using visuals for
general audiences to better comprehend.
• Interactive data stories include innovative visualizations with narrative.

Social media card or factoid card

• A quick snapshot (1-2 charts) of a key insight which could be taken from a
report or created with data completely separate from any other product.
Interactive visual for social media
• An interactive visual, collection of animated visuals, or data story that
allows users to interact with the chart, hover over parts of the visual to get
exact data values and information and explore the various elements
 How are data visualizations disseminated?

Report (i.e. Gender Factbook)

• Official statistical output with mostly text and analysis supported by visuals

Complementary executive summary/visual brief

• 1-2 pages long summarizing key visuals and insights from report

Infographic
• An eye-catching collection of images, icons, charts with minimal text

Written blog or data story

• An article that tells a narrative/story around the data using visuals for
general audiences to better comprehend.
• Interactive data stories include innovative visualizations with narrative.

Social media card or factoid card

• A quick snapshot (1-2 charts) of a key insight which could be taken from a
report or created with data completely separate from any other product.
Interactive visual for social media
• An interactive visual, collection of animated visuals, or data story that
allows users to interact with the chart, hover over parts of the visual to get
exact data values and information and explore the various elements
 How are data visualizations disseminated?

Report (i.e. Gender Factbook)

• Official statistical output with mostly text and analysis supported by visuals

Complementary executive summary/visual brief

• 1-2 pages long summarizing key visuals and insights from report

Infographic
• An eye-catching collection of images, icons, charts with minimal text

Written blog or data story

• An article that tells a narrative/story around the data using visuals for
general audiences to better comprehend.
• Interactive data stories include innovative visualizations with narrative.

Social media card or factoid card

• A quick snapshot (1-2 charts) of a key insight which could be taken from a
report or created with data completely separate from any other product.
Interactive visual for social media
• An interactive visual, collection of animated visuals, or data story that
allows users to interact with the chart, hover over parts of the visual to get
exact data values and information and explore the various elements
 How are data visualizations disseminated?

Report (i.e. Gender Factbook)

• Official statistical output with mostly text and analysis supported by visuals

Complementary executive summary/visual brief

• 1-2 pages long summarizing key visuals and insights from report

Infographic
• An eye-catching collection of images, icons, charts with minimal text

Written blog or data story

• An article that tells a narrative/story around the data using visuals for
general audiences to better comprehend.
• Interactive data stories include innovative visualizations with narrative.

Social media card or factoid card

• A quick snapshot (1-2 charts) of a key insight which could be taken from a
report or created with data completely separate from any other product.
Interactive visual for social media
• An interactive visual, collection of animated visuals, or data story that
allows users to interact with the chart, hover over parts of the visual to get
exact data values and information and explore the various elements
 How are data visualizations disseminated?

Report (i.e. Gender Factbook)

• Official statistical output with mostly text and analysis supported by visuals

Complementary executive summary/visual brief

• 1-2 pages long summarizing key visuals and insights from report

Infographic
• An eye-catching collection of images, icons, charts with minimal text

Written blog or data story

• An article that tells a narrative/story around the data using visuals for
general audiences to better comprehend.
• Interactive data stories include innovative visualizations with narrative.

Social media card or factoid card

• A quick snapshot (1-2 charts) of a key insight which could be taken from a
report or created with data completely separate from any other product.
Interactive visual for social media
• An interactive visual, collection of animated visuals, or data story that
allows users to interact with the chart, hover over parts of the visual to get
exact data values and information and explore the various elements
 For whom are data visualizations produced?
➢ The target audience for communications regarding gender statistics is policymakers and funders who
are key decision-makers as well as the people who influence them like the general public (voters),
media, advocacy organizations etc.

Right data
Right format (topic, priority
(meets users’ issues,
needs) statistically
sound, quality)

Right users
(reaches intended users)
Mapping data outputs to audiences

Data analysts,
researchers, academia
•Complete granular raw data,
Development practitioners, questionnaires, codebooks, etc.
gender specialists in •Dedicated/comprehensive
government, NGOs databases of microdata and
Media, policymakers •Visuals of both summary and processed gender statistics
disaggregated data •Advanced visualizations often
•Summary tables, charts, trends,
visualizations, short stories •Analysis of results including using statistical concepts

General public trends over time in standard

•Press release or short factsheet reports, metadata
•Key figures and visualizations
•Short factsheet or infographic
•Social media graphic or blog
High level of detail

Moderate level of detail

Lower level of detail

 PRACTICE EXERCISE
➢ Go to Sheet named "Exercise 1" in the Excel file "Training Dataset Day 1".
➢ Review the dataset and answer the following questions:

1. What are the key or most useful insights from the data?
• Which disaggregations do you wish to communicate in the visualization?

What We Do
2. Who is/are the audience(s) you are communicating to?

3. Through which type of data output would you disseminate the data visualization?
• If you have multiple audiences, will there be more than one data output? Which types?

4. What type of potential actions can be taken based on the insights in the data visualization?

➢ Time for exercise: 15 minutes

➢ Write your answers in the Sheet named "Ex1 Response".
STRENGTHENING
GENDER
STATISTICS

DATA VISUALIZATION
TRAINING MODULE 2:
DATA VISUALIZATION
PRINCIPLES & CONCEPTS
Contents
1. Data visualization overall principles
2. Data visualization steps
3. Data visualization tips and concepts by chart type
1. Overall data
visualization
principles
Data visualization improves communication
GUIDING PRINCIPLES BASIC RECOMMENDATIONS
1. Understand the information/data and 1. Highlight 1-2 key messages per visual.
prioritize the information for sharing.
2. Explore graph and chart design choices beyond
2. Choose the correct software/platform and the default options.
chart type for effective visuals designs keeping
in mind the audience, data, output type. 3. Recognize the right use of visual characteristics
especially color as it is incredibly powerful to
3. Use effective data visualization concepts and depict meaning.
characteristics (color, shape, size, pattern, etc.)
to maximize the impact of the data. 4. Reduce the clutter and keep only essential
elements (no duplicative visual characteristics)
4. Communicate data meaning clearly, quickly
and ethically. 5. Never mislead the audience/manipulate visual.

5. Always integrate text elements with the 6. Use annotations, minimize jargon, acronyms,
graphs and images to tell a story (consider and technical terms, and choose a font that is
infographics). easy to read.

BENEFITS
Decision-makers more quickly absorb gender-related insights from surveys
which in turn facilitates evidence-based policy-making
 1. Understand and show the data
➢ To create a great visualization, you must understand the key insights and how to let the data shine
through the visual.
➢ Auto-generated charts will not automatically or adequately highlight the important aspects of the data.
➢ You must explore the data first to understand which insights to relay to audience.
• Do not include too many variables or too much information. You do not have to relay all the data.
Other data can go in a table in annex.
• Include helpful annotations.
 2. Select the right type of graph
➢ Match the graph type to the audience’s level of data/statistics comprehension.

Bar Line Area chart Geographical map

• Bar and column chart (grouped or
stacked), line chart, area chart,
Non-expert
audiences geographic map, range plot, pie chart

• Heatmap, tree map, histogram, Dot plot Histogram Heat map Treemap
distribution chart, dot plot, complex
Medium
expertise tables

Scatter plot Box plot Gantt chart

• Scatter plots, boxplots, Sankey or
Expert alluvial diagram, Gantt chart
audiences
 2. Select the right type of graph
➢ Match the graph to the type of data and the message you want readers to gain.
• Categorical (bar, column, treemap), continuous (scatterplot, heatmap), time series (line, area chart, heatmap).

Bar Line Area chart Geographical map

• Showing how values are similar

Comparison or different, highlighting gaps
Stacked bar and column charts Area chart Pie chart Treemap
• Parts of a total that usually add
Composition up to 100%
Scatter plot Treemap Gantt chart
• How variables relate to one
Relationship another

Scatter plot Histogram Box plot Dot plot

• Where values fall within the
Distribution dataset, identifying outliers
 3. Select the data visualization characteristics
➢ Our eyes detect changes or highlights through visual perception of the following characteristics.
 4. Make graphs clear and clean
➢ Declutter but don’t oversimplify chart!
• When possible, reduce numbers to simplest
form.
• Label clearly, specify units, use a legend when
necessary.
• Remove duplicated characteristics (gridlines vs.
numeric labels; color vs. data labels).
• Avoid 3D graphics and shadows. They distort
data and it’s not easy to visually see differences
between the chart cylinders, columns etc.
• Break the chart into smaller multiples if the
original is too cluttered.
 5. Integrate text elements or annotations
➢ Text can highlight a particular data point or guide the audience through the content.
• This should be in addition to the supplementary text/context in a report that refers to the visual.
 6. Choose colors and fonts wisely
➢ Colors should aid understanding, not distract readers.
• Try a maximum of 3 colors (unless categorically required).
• Stick with the same fonts and colors consistently.
• Do not use “familiar colors in surprising ways” i.e. red for
good and green for bad.
• Check whether the colors and contrast work for color
blindness and for printing black & white - you may not see
the data labels on the bars.

➢ Gender data and color

• Do not use stereotypical male and female colors (blue/pink).
• Stick with the same colors for variable groups consistently
throughout a product or report.
• Female and Male or Urban, Rural, Total.
 6. Choose colors and fonts wisely
➢ Colors should aid understanding, not distract
readers.
• Sequential or gradual color scales are
usually for continuous variables. In some
cases, they work for stacked bar/column
charts.
• Diverging color scales are usually for
gender gaps where the values are moving
away from 0 in either direction.
• Categorical color scales should be used for
multiple categories or variables, but never
be used for a different color for each
observation within a category.
• Missing data or “other” categories should
be a neutral color like grey or grey
patterned for maps.
 GROUP DISCUSSION
➢ Discussion – which colors
are used for male and
female in this report?
➢ Are they consistent?

➢ Is there anything you

would change about the
colors of these visuals?
 GROUP DISCUSSION

F: Orange
M: Blue

F: Pink
M: Green
 GROUP DISCUSSION
➢ Discussion – which colors Color or race
are used for male and
female in this report?
➢ Are they consistent? Total
Men
➢ Is there anything you
Women
would change about the
colors of these visuals? White Black or brown

Population groups by income

Total

20% with the lowest

income

20% with the highest

income
Men Women
 GROUP DISCUSSION

Do not reuse the Female & Male

colors for other categories!
These two colors should remain distinct
in order to not confuse the reader
throughout the report or infographic.
 7. Avoid misleading the audience
➢ Do not allow for jumps in axis labels.
• Horizontal axis must have consistently or equivalently spaced years. For example, one cannot skip
from 2000 to 2010 and then yearly.
 7. Avoid misleading the audience
➢ Do not allow for jumps in axis labels.
• If vertical axis does not start at 0 – MUST have a symbol to denote the break in axis.

➢ Do not add values in a stacked bar or column or in a pie chart that do not total 100% For example,
stacking female, male, and total together.
➢ Don not use inconsistent scales.
 8. Highlighting the gender-relevant insights
➢ NSOs' comparative advantage regarding statistics → highlight detailed levels of disaggregation that
international organizations don’t calculate or disseminate.
1. Use multiple disaggregations in the same charts whenever possible.
➢ Aesthetically more pleasing than just 2 bars for male and female and provide more gender-
relevant insights that tell a more nuanced story.
2. Use subnational region maps.
➢ International databases cannot showcase this subnational level as it’s not internationally
comparable.
 8. Highlighting the gender-relevant insights
➢ Aim to highlight the gender gap rather than the distribution of female and male values across a different
category (age, sector, type of employment, etc.).

Showcasing the distribution of female/male employment across sectors Showcasing the gender gap

Key insight: Most women and men work in the agricultural sector Key insight: Gender gap in employment is biggest in the industrial sector
2. Data visualization
steps
Steps for creating a data visualization

Explore/try
visualization
options Three-fourths of creating a visualization is add
Optional: cleaning
Design, format, Publish and/or
Upload data and finalizeand formatting the data in the proper additional
download way to:
visualization visualization annotations
as needed
Adjust data match the required data structure, inputs,
structure
and features of the intended visualization type

highlight the right message and insights

enable easy formatting and annotation

Data visualization breakdown

25%
Formatting &
75% annotating
Three-fourths of creating a visualization is cleaning
the visual
Properly and formatting the data in the proper way to:
cleaning,
match the required data structure, inputs,
transposing and features of the intended visualization type
& preparing
the data highlight the right message and insights

enable easy formatting and annotation

 Data structure
➢ The type of visualizations you can create depends on the structure of the data tables (columns and
rows).
➢ See below two different structures of the same data points.

A. Grouped by gender B. Grouped by type of work

Gender Unpaid Domestic Work Unpaid Care Work Type of Work Female Male

Unpaid Domestic Work 15.9 4.5

Female 15.9 3.4

Unpaid Care Work 3.4 .6

Male 4.5 .6
 Data structure
➢ The type of visualizations you can create depend on the structure of the data tables (columns and rows).
➢ See below two different structures of the same data points with grouped bar chart examples.

A. Grouped by gender B. Grouped by type of work

Gender Unpaid Domestic Work Unpaid Care Work Type of Work Female Male

Unpaid Domestic Work 15.9 4.5

Female 15.9 3.4

Unpaid Care Work 3.4 .6

Male 4.5 .6

Difference/gap between the types of unpaid work for a given gender. Difference/gap between men’s and women’s time spent on a given
type of unpaid work.
3. Data visualization
tips and concepts by
chart type
 What We Do

COMPARISON – BAR AND COLUMN CHARTS

 Bar and column charts
➢ Simplest form of charts which allow for easy interpretation by non-expert audiences.
➢ Bar charts allow the labels to be legible on the left side whereas column charts the labels might be at a
45 to 90 degree angle.
➢ If not sex-disaggregated (i.e. adolescent fertility), keep all bar colors the same and highlight only the
bar with the insight you’re imparting on audience in separate color.
➢ Order bars by value and not alphabetically (for example geographical administrative regions).
 Bar and column charts
➢ If highlighting female vs male values, group the bars or columns by
non-gender category and use only two colors – one for each gender.
➢ Use gridlines or data labels – don’t use both.
➢ Data labels are acceptable for up to 10-15 columns/bars
➢ Consider using icons when there are few bars. For example, only two
bars (Female/Male) or four (i.e. Female/Male with Urban/Rural
disaggregation).
➢ Sometimes column charts are used to show survey years (as an
alternative to line charts).
 What We Do

COMPARISON – MAPS
 Maps
➢ Maps will provide nuance to the national statistics by easily identifying the
discrepancies within the country regionally.
• These regional data are often not collected by international
organizations but are most crucial to policymakers in order to pinpoint
policies and interventions that work for each region.
➢ Use the same color scale and ranges when using side by side maps to
compare female and male values.
➢ To avoid side by side maps, map the gender gap rather than female and
male values.
• Make sure to use a diverging scale with the gender gap data.
• Rename the legend for the direction of the gap whether the female
value is relatively higher or lower than the male value or vice versa.
• Negative numbers will confuse the reader. Where possible don’t show
the negative numbers.
 GROUP DISCUSISON

➢ Discussion – Is it easy to see the

main insights? Easy to see the
gender gap?

➢ Is there anything you would

change about the colors of
these visuals?
 GROUP DISCUSISON
➢ Discussion – what is the
takeaway from the map?

➢ Is there anything you would

change about the colors of
these visuals?

➢ Potential actions:
• Financial literacy classes in
school.
• Support to women owned
business.
• Simplify the loan approval
process.
 What We Do

TRENDS OVER TIME – LINE CHARTS

 Line charts
➢ Used primarily for time series with year on x-axis.
• X-axis could also be used for ordinal categories like age ranges, hours in a day, wealth quintile.
➢ Policymakers and researchers are most interested in seeing progress over the years for narrowing
gender gaps, so where possible, provide time-series for indicators.
➢ Differentiate lines first by color, then by pattern (dashed or dotted line).
➢ Especially important in time series to highlight years that may show stark contrast in data before and
after that year.
• Year in which methodology/calculation of indicators according the ICLS -19 were implemented in
data collection as there may be a stark contrast in the data.
• Year of historical event or relevant legislation enacted.
 Line charts
➢ There should be no more than 4-6 lines in a line chart.
➢ Use small multiple when there are too many lines. This is also called a grid of charts and you can
easily show male and female values for several disaggregations.
 Line charts
➢ Show female and male rates while also easily visualizing the gap.
• Shade the gap area between the female and male lines to make the gender gap more visible.
➢ Annotate and add highlight ranges for context.
 What We Do

COMPOSITION - PARTS OF A WHOLE

 Stacked bar and column charts
➢ Used for depicting parts of a whole/total typically 100%
➢ Alternative to the grouped bar or column chart to save space.
➢ Most commonly used to stack:
• female and male.
• age ranges.
• categories of employment (part time vs full time, sectors, etc.).
• reasons for not being in labor force.
• categories of unpaid work.
 Stacked bar and stacked area charts

➢ Decision to group by gender or by

disaggregation depends on the
message being highlighted.
Highlighting gender gap vs
highlighting distribution or
comparison within disaggregation
category.

➢ Stacked area charts convey the

same things as stacked bar or
column chart but with the time
series dimension.
 GROUP DISCUSISON

➢ Discussion – Is it easy to see the main insights? Easy to see the gender gap? Is there anything you
would change about the colors of these visuals?
 What We Do

COMPARISON OF GAPS – RANGE PLOTS

 Range plots
➢ Great for highlighting gender gaps very clearly, especially
when there are a lot of categories or disaggregations (age,
location, region, etc.).
➢ Show female and male rates using a colored dot
➢ Show gender gap using the length of the shaded/colored
line between the two dots.
 Range plots
➢ Can add the data labels for both the length of the line (gap) and the dots (female and male values).
➢ The values will always be female vs. male will NOT show totals.
➢ You could potentially show range plot between two years rather than between female and male values.
➢ The range plot should always be ordered by value (usually female) so it’s easy to follow.
 What We Do

COMPARISON/TIME TRENDS – HEATMAPS

 Heatmaps
➢ Convey a lot of information with multiple
disaggregations in one chart.
➢ Can have one row of female values and one
row of male values for comparison.
➢ Can visualize the gender gap for a
disaggregation category with many
categories (quintile, types of ownership,
age ranges).
 What We Do

DISTRIBUTION – DOT PLOTS

 Distribution chart/Dot Plot
➢ Shows distribution within the dataset/category
➢ Country value compared to regional values.
➢ Capital city value compared to administrative
region values.
➢ Distribution of all household values (if using
microdata).
➢ Can show vertically or horizontally and color
code by region or other types of disaggregation.
 What We Do

PRACTICE EXERCISES
 Exercise – suggest changes to the set of visuals
➢ See Sheet named "Exercise 2" in the Excel file "Training Dataset Day 1" for an example of a set of visuals
that are part of the same report.
➢ Write a list of suggested changes that you would make to this set of visuals to ensure that they are
consistent and following best practices in data visualization principles.

➢ For best practices in data visualization refer to the "Data Visualization Tip Sheet" handout or Word file.
➢ There should be consistency across the whole report in:
• colors for the same categories;
• font type and size;
• title structure and punctuation;
• phrasing of categories.

➢ Time for exercise: 20 minutes.

➢ Write your answers in the Sheet named "Ex2 Response".
ANNEX
Financial Times Visual
Vocabulary
Range plot examples
STRENGTHENING
GENDER
STATISTICS

DATA VISUALIZATION
TRAINING MODULE 3:
EXCEL CHARTS AND DATA
ADJUSTMENTS
Contents
1. Recap of data visualization steps and breakdown
2. Using data visualization principles to adjust data and chart
elements in Excel
- Grouping or ordering data, chart elements,
and legend for the right data presentation
- Decluttering the chart
- Appropriate sorting and use of colors
 Specific actions for adjusting chart elements

1. Using the right data structure or grouping to highlight the key

gender-relevant insight (slides 9-16).
2. Reordering chart elements to ensure visually female values
are before the male values (slides 18-25)
3. Adjusting data structure for grouping multiple
disaggregations in one chart (slides 27-38)
4. Decluttering the visual (slides 41-58)
• Removing 3D elements (slides 41-43)
• Removing colored background (slides 44-48)
• Removing gridlines (slides 49-50)
• Removing axis title and labels (slides 51-52)
• Formatting data labels (slides 53-56)
• Sorting values in chart (slides 57-58)
5. Adjusting the colors of values in a chart (slides 60-64)
6. Removing the legend (slide 65)
1. Recap of data
visualization steps
and breakdown
Recap: steps for creating a data visualization

Explore/try
visualization
options Three-fourths of creating a visualization is add
Optional: cleaning
Design, format, Publish and/or
Upload data and finalizeand formatting the data in the proper additional
download way to:
visualization visualization annotations
as needed
Adjust data match the required data structure, inputs,
structure
and features of the intended visualization type

highlight the right message and insights

enable easy formatting and annotation

Recap: data visualization breakdown

25%
Formatting &
75% annotating
Three-fourths of creating a visualization is cleaning
the visual
Properly and formatting the data in the proper way to:
cleaning,
match the required data structure, inputs,
transposing and features of the intended visualization type
& preparing
the data highlight the right message and insights

enable easy formatting and annotation

Recap: data visualization breakdown
2. Using data visualization
principles in Excel to adjust
data and chart elements
 USING THEWhat
RIGHTWe DoSTRUCTURE
DATA
OR GROUPING TO HIGHLIGHT THE KEY
GENDER-RELEVANT INSIGHT
 Adjusting data grouping
➢ Time use data (% of time spent of a 24-hour day) – values are grouped by type of unpaid work.
➢ You decide to select a stacked bar chart to visualize parts of a whole.
• What part/percentage of the whole day do these work activities comprise for women and men?

Grouped by type of work

Type of Work Female Male

Unpaid Domestic Work 15.9 4.5

Unpaid Care Work 3.4 .6

 Adjusting data grouping
➢ Highlight the data range and click the “Insert” tab in the Excel ribbon (toolbar menu at the top).
➢ Under the "Charts" section click “Recommended Charts” and click the stacked bar chart and click “OK”.
2. 3.

1.

4.

5.
 Incorrect data structure
➢ This visual is grouping values by type of work (row) and stacking female and male values (columns).
➢ The data are not structured properly for this stacked bar chart to show what part/percentage of the
whole day these unpaid work activities comprise for women and men.

Chart Title Adjustment required!

This visual should instead be stacking types
of unpaid work to show the % of time spent
Unpaid Care Work in a day by each gender on the combined
unpaid domestic and care work activities.

Unpaid Domestic Work Type of Work Female Male

Unpaid Domestic 15.9 4.5

0 2 4 6 8 10 12 14 16 18
Work
#REF! Female Unpaid Care Work 3.4 .6
 Adjustment option 1: transpose data in tool
➢ First, always try transposing the data in the visualization software.
• Click into the chart and then click on the "Chart Design" tab in the Excel ribbon (top toolbar menu). If
there is no "Chart Design" tab try clicking into the chart again because this option will only show up
when the chart is selected.
• Then click "Switch Row/Column" in the Excel ribbon (top toolbar menu).
2. 3.

1.
 Adjustment option 1: transpose data in tool
➢ First, always try transposing the data in the visualization software.
➢ You should see that the visual has changed. The visual is grouped by gender now. This does not
change the data table in the Excel spreadsheet. It only changes the chart orientation.

Chart Title

Female

Male

0 5 10 15 20 25

Unpaid Domestic Work Unpaid Care Work

Easier to see that women spent much

higher percentages of their day on
unpaid work than men.
 Adjustment option 2: change visual type
➢ If you are not set on the chart type, switch to a visualization that better fits the original data structure.
• Select a different chart type icon that you wish to try (i.e. stacked column, grouped bars/columns).

Alternative visuals
30

20
Emphasis on
gender gap within
10
a given type of
0
Female Male
unpaid work.
Intended visual Unpaid Domestic Work Unpaid Care Work

Female
Unpaid Care Work Emphasis on
Male
gender gap within
Unpaid Domestic Work
0 10 20 30 a given type of
0 5 10 15 20 unpaid work.
Unpaid Domestic Work Unpaid Care Work
Female Male
Emphasis on gap or difference between the
types of unpaid work for a given gender. 20
Emphasis on gap or
10
difference between
0 the types of unpaid
Female Male
work for a given
Unpaid Domestic Work Unpaid Care Work gender.
Adjustment option 2: change visual type
How do you choose which Alternative visuals
- Correct message
alternative visual to use? 20
- Data labels and
10 color key needed
• Is the visual still emphasizing or 0 - Make chart area
highlighting the right Female Male taller and bars
Unpaid Domestic Work Unpaid Care Work slightly thinner
message/insight?
• Do you have to adjust the chart
Unpaid Care Work
area or the bar or column - Correct message
- Minimal aesthetic
height/width? Unpaid Domestic Work
changes or
• Do you have to add data labels, 0 5 10 15 20 ordering needed
Female Male
a color key, or gridlines?
• Do you have to sort, regroup, or 20 - Must transpose
re-order the bars or columns? 10 to highlight the
correct insight
0
Female Male
- Data labels or
Unpaid Domestic Work Unpaid Care Work gridlines needed
- Make chart area
taller and bars
slightly thinner
 TASK 1
➢ Go to Sheet named "Task 1" in the Excel file "Training Dataset Day 1".
➢ Use either of the options demonstrated (transposing data or changing chart type) to adjust the chart in
the sheet so that it highlights the gender gap.
 What
REORDERING We ELEMENTS
CHART Do TO
ENSURE THAT VISUALLY THE FEMALE VALUES
ARE BEFORE THE MALE VALUES
 Adjusting the order/display of genders
➢ Female values should always be first in the chart either above the male values or to the left of the male
values.
➢ In this chart, the male values show up first, above the female values.
• The data table must be reordered to make the change show up in the chart.

Chart Title

Male

Visually, "Male" is
ordered above or Female

before "Female".
0 5 10 15 20 25

Unpaid Domestic Work Unpaid Care Work

 Adding a blank column
➢ Edit the data table by inserting a column in between the "Female" and "Male" column.
• Click the letter C at the top of the "Male" column to highlight the entire column C for "Male".
• Right click the highlighted column and select "Insert".
• There should now be a blank column in between the "Female" and "Male" columns.
 Copying and pasting tips
➢ Right click the highlighted cells and select "Copy" then right click the top cell in the empty column and
select "Paste" which might look like a clipboard and paper symbol instead of the word "Paste".
➢ Shortcut: CTRL+C for copy and CTRL+V for paste.
• After highlighting the cells, hold the CTRL and the C key at the same time to copy the highlighted
values.
• Then click on the top/first cell in the empty column and hold the CTRL and the V key at the same
time to paste the values.
 Copying and pasting data into blank column
➢ Rearrange the columns so that the male values are to the left of the female values in the table.
• Highlight the cells in the "Male" column by clicking and holding the first cell where it says "Male" and
dragging down to the last row. All three rows should be highlighted.
• Then copy and paste the male data into the empty middle column.
• Male values should be duplicated now in both the table and the chart.

Chart Title

Male

Male

Female

0 5 10 15 20 25

Unpaid Domestic Work Unpaid Care Work

 Copying and replacing data in column
➢ Rearrange the columns so that the male data are to the left of the female data in the table.
• Copy the "Female" column and replace the "Male" column furthest to the right.
• Female values should be duplicated now in both the table and the chart.

Chart Title

Female

Male

Female

0 5 10 15 20 25

Unpaid Domestic Work Unpaid Care Work

 Deleting the duplicated column
➢ Rearrange the columns so that the male data are to the left of the female data in the table.
• Delete the first column of data which is column B for "Female" by right clicking the top of the column
where it says "B" and selecting "Delete".
 Adjusting the order/display of genders
➢ You may receive this pop-up warning message. Click "OK".
➢ The data table will now only have two columns and the chart will show "Female" above "Male".

Chart Title

Female

Male

0 5 10 15 20 25

Unpaid Domestic Work Unpaid Care Work

 TASK 2
➢ Go to Sheet named "Task 2" in the Excel file "Training Dataset Day 1".
➢ Correct the Female/Male ordering in the chart in the sheet (preview below).
 What
ADJUSTING DATA We DoFOR GROUPING
STRUCTURE
MULTIPLE DISAGGREGATIONS
IN ONE CHART
 Adjusting grouping of multiple disaggregations
➢ The table is showing asset ownership (% of females or males who have sole, joint or both sole and joint
ownership of land).
➢ The data are structured as seen below by gender, type of ownership and location.

Ownership Female Male

Capital City
Sole ownership 21.0 79.0
Joint ownership
33.4 66.6
Both sole and joint ownership 21.4 78.6
Urban

Sole ownership 28.2 71.8

Joint ownership
15.7 84.3
Both sole and joint ownership
24.9 75.1
Rural
Sole ownership 22.2 77.8
Joint ownership 9.5 90.5
Both sole and joint ownership 20.2 79.8
 Adding a clustered column chart
➢ Highlight the data range and click the “Insert” tab in the Excel ribbon (toolbar menu at the top).
➢ Under the "Charts" section click “Recommended Charts” and click the clustered column chart and click “OK”.
2.
3.

1.

4.

5.
 Default chart view of multiple disaggregations
➢ The visual does not register the location grouping in this data structure and thinks it is part of the series.
➢ The data must be restructured so that the Excel visualization tool can cluster the column charts by
location.
Chart Title
100.0

90.0

80.0

70.0

60.0

50.0

40.0

30.0

20.0

10.0

0.0
Sole ownership
Joint
Bothownership
sole and joint ownership Sole ownership
Joint
Bothownership
sole and joint ownership Sole ownership
Joint
Bothownership
sole and joint ownership

Women Men
 Inserting a new column
➢ To restructure the data, first insert a new column to the left of the "Ownership" column (currently column
A) by right clicking on the top of Column A and selecting “Insert".
➢ There should now be a blank first column.
 Cutting and pasting data into blank column
➢ Cut the location categories and paste them into the first column in the first row of data for that section.
• You must cut each category and paste it individually. Repeat for each category that must be moved.
▪ Shortcut for the cut function is CTRL+X (holding CTRL and X at the same time after the intended
cell has been highlighted).
• In this example, the name of the location should be in the same row as the
“Sole ownership” data.
 Opening the select data source window
➢ Right click on the chart and click “Select Data”.
➢ The “Select Data Source” box will pop up.
 Selecting the new range of data in data source
➢ The previous data range will be denoted with a dashed green border.
➢ Select the full range of data including column A and hit the "Enter" button on the keyboard or click the
"OK" button in the popup box.
 Adjusting grouping of multiple disaggregations
➢ Your visualization should now be grouped by the additional “location” disaggregation but there is a little
too much spacing in between the groups.

Chart Title
100.0

90.0

80.0

70.0

60.0

50.0

40.0

30.0

20.0

10.0

0.0
Sole Joint Both sole Sole Joint Both sole Sole Joint Both sole
ownership ownership and joint ownership ownership and joint ownership ownership and joint
ownership ownership ownership
Capital City Urban Rural

Women Men
 Removing extra space between groupings
➢ Highlight the empty rows between the groupings by clicking on the grey number associated with the row.
➢ The rows can be highlighted and deleted separately or at the same time. To highlight and delete at
the same time, highlight the first row as indicated above, then hold down the CTRL button and
highlight the second empty row you plan to delete (while still holding the CTRL button) before
moving to the next step.
 Removing extra space between groupings
➢ Once the rows you want are highlighted, right click the row and select “Delete”.
• If you accidentally highlight the wrong row, before you delete the row, click anywhere in the Excel
sheet to unhighlight, then start again with highlighting the right row as per the previous slide’s
instructions.
➢ If you are deleting each row separately, highlight the next row and repeat the process.
 Adjusting grouping of multiple disaggregations
➢ Your visualization should now be grouped by the additional “location” disaggregation with no additional
space in between the groups.

Chart Title
100.0
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0
Sole ownership Joint ownership Both sole and Sole ownership Joint ownership Both sole and Sole ownership Joint ownership Both sole and
joint ownership joint ownership joint ownership
Capital City Urban Rural

Women Men
 TASK 3
➢ Go to Sheet named "Task 3" in the Excel file "Training Dataset Day 1".
➢ Adjust the chart in the sheet to have the multiple disaggregations neatly grouped within the chart rather
than individually listing all of the different combinations.
 GROUP DISCUSSION
➢ Below is a visual that you have received in a report and must apply data visualization principles to fix it.
➢ Discussion – What are the immediate issues with this visual? What changes would you make?

FEMALE TO MALE EMPLOYMENT RATIO

Household with children Household without children Couple without children
Couple with children Extended Family Overall

0.64
0.60 0.61 0.62 0.61

0.53

Female to Male Employment Ratio

 What We Do
DECLUTTERING:
REMOVING 3-DIMENSIONAL ELEMENTS
 Selecting a 2-dimensional chart
➢ To remove the 3-Dimensional effect, click on the chart and under the “Chart Design” tab select “Change
Chart Type” and select the 2-Dimensional clustered columns and click “OK”.
2. 3.

1.
4.

5.
 Decluttering visuals
➢ The visualization should now be 2-Dimensional.

FEMALE TO MALE EMPLOYMENT RATIO

Household with children Household without children Couple without children
Couple with children Extended Family Overall

0.64
0.61 0.62 0.61
0.60

0.53

Female to Male Employment Ratio

 What We Do
DECLUTTERING:
REMOVING COLORED BACKGROUND
 Selecting a new chart style
➢ You can remove the background either by changing the whole chart style or by changing the color of the
background.
➢ To change the chart style, click on the chart and click on the paintbrush icon to the right and select a new
chart style or go to the "Chart Design" tab and select a new chart style from the designs.
 New chart style with clear background.
➢ The visualization should now have a clear background.

Female to male employment ratio

0.70
0.64
0.61 0.62 0.61
0.60
0.60
0.53

0.50

0.40

0.30

0.20

0.10

0.00
Female to Male Employment Ratio

Household with children Household without children Couple without children

Couple with children Extended Family Overall
 Changing the background color of chart
➢ To just change the background color and none of the other chart elements, click the chart and go to the
“Format Chart Area” panel that pops up on the right side of the screen. Under the paint icon under the
“Fill” section, click on the "Color" dropdown and select white.
 Changing the chart style vs. background color
➢ The left is the visual after changing the chart style while the right is the visual after changing the
background back to white.

Female to male employment ratio FEMALE TO MALE EMPLOYMENT RATIO

0.70 0.64 Household with children Household without children Couple without children
0.60 0.61 0.62 0.61
Couple with children Extended Family Overall
0.60
0.53
0.64
0.50 0.60 0.61 0.62 0.61
0.53
0.40

0.30

0.20

0.10

0.00
Female to Male Employment Ratio

Household with children Household without children Couple without children Female to Male Employment Ratio
Couple with children Extended Family Overall
 What We Do
DECLUTTERING:
REMOVING GRIDLINES
 Removing gridlines
➢ You can remove gridlines in multiple ways. You can click on the gridlines in the chart (the blue circles
confirm that the gridlines are selected and not another chart element) and then hit the “Delete” key.
➢ Or you can click on the chart and then the “Chart Elements” plus symbol and uncheck the box next to
"Gridlines".
 What We Do
DECLUTTERING:
REMOVING AXIS TITLE AND LABELS
 Removing the axis title and axis labels
➢ There are now duplications of elements. The y-axis is duplicating the data labels while the x-axis is
duplicating the chart title. They can be removed in the same way as the gridlines either by clicking on the
chart element and hitting the “Delete” key.
➢ Or you can click on the chart and then the “Chart Elements” plus symbol and uncheck the box next to
"Axes".
FEMALE TO MALE EMPLOYMENT RATIO
0.70 0.64 0.62 0.61 0.61 0.60
0.60 0.53
0.50
0.40
0.30
0.20
0.10
0.00
Female to Male Employment Ratio

Household with children Extended Family

Overall Couple with children
Household without children Couple without children
 What We Do
DECLUTTERING:
REFORMATTING DATA LABELS
 Moving data labels inside the bars
➢ The data labels are better placed inside the columns. They can be changed by clicking on the chart and
then on the “Chart Elements” plus symbol and clicking on the “Data Labels” option.
➢ Select “Inside End” or “Inside Base” for the "Data Labels".
 Opening the “Format Data Labels” panel
➢ The data labels will appear inside the bars, but they cannot be seen well. The format should be changed.
➢ Data labels can be edited either by clicking a data label and going to “Format Data Labels” panel on the
right side or by clicking the “Chart Elements” plus symbol, going to “Data Labels”, and then “More
Options..” which should also open up the “Format Data Labels” panel.
 Decluttering visuals
➢ In the “Format Data Labels” panel, click the “Text Options” then select a better contrasting color against
the colors of the columns from the “Color” dropdown. Click into each data label and repeat color changes.
 What We Do
SORTING VALUES IN A CHART
 Sorting values within the chart
➢ To sort the columns, bars, or other elements for other types of visuals, the sorting must happen in the
original data table.
➢ Click in the column for the variable which you wish to sort. Then under the “Home” tab in the Excel
ribbon (top toolbar menu, click on the “Sort & Filter” button in the “Editing” section. Then click the “Sort
Z to A” or “Sort A to Z” option depending on which direction you require the values to flow.
2. 3.

4.
1.
 TASK 4
➢ Go to Sheet named "Task 4" in the Excel file "Training Dataset Day 1".
➢ Adjust the chart in the sheet to remove any unnecessary elements (3D, gridlines, axes, etc.).
➢ Adjust the chart in the sheet to sort the bars in order of value.

Chart Title
66.8
REGION G 44.2
89.3
48.7
REGION F 41.6
55.7
49.5
REGION E 48.1
50.9
44.2
REGION D 40.7
47.6
47.6
REGION C 46.2
48.9
43.4
REGION B 41.6
45.1
51.6
REGION A 52.6
50.5

0 10 20 30 40 50 60 70 80 90

Internet Access Total Internet Access Male Internet Access Female

 What We Do
ADJUSTING COLORS OF VALUES IN A CHART
 Adjusting colors
➢ The data shown are categorical in the column chart, but the color is not encoding any particular insight.
There should only be one color for all of the columns.
➢ When customizing colors you can i) select one of the colors already in the pop-up box or ii) enter a 6-
letter-number combination called a HEX code.
• For this example, let’s use the HEX code #440E5F which produces this color .
• If you do not have the option to enter a HEX code you may enter the corresponding RGB code with
three numbers. For this color the RGB code is (68, 14, 95). Each of the three numbers corresponds to
the three letters R (for Red), G (for Green, and B (for Blue).

FEMALE TO MALE EMPLOYMENT RATIO

0.64
0.62 0.61 0.61 0.60
0.53

Household with children Extended Family Overall

Couple with children Household without children Couple without children
 Selecting hex codes
➢ Each color has a code made up of 6 units that are a combination of letters and numbers. If you want to
consistently use the same color, write down the code for that color to use in multiple visualizations.
• All HEX codes have corresponding RGB codes in case you do not have the option to enter HEX codes.
➢ HEX codes can be found at htmlcolorcodes.com and https://www.computerhope.com/htmcolor.htm
• To find the corresponding RGB code of your HEX code, use https://htmlcolorcodes.com/hex-to-rgb/
➢ You can use the HEX codes that correspond to the colors in the National Statistical Office’s branding.
 Entering a hex code in the “Format Data Series”
➢ Click on a column and go to the “Format Data Series” panel that pops up in the right side of the screen.
➢ Click the paint icon and go to the color dropdown.
➢ Instead of choosing a color from the pop-up, select “More colors” this will prompt the "Colors" window
to pop up. Go to the “Custom” tab.
➢ At the bottom of the pop-up there is a fields titled “Hex”, "Red", "Green", and "Blue". Replace the HEX or
RGB code of the current color with the NSO’s branding color or the agreed upon color for the
report/visual. To the right it shows a preview of the new color against the old color. Then click “OK”.
 Adjusting colors
➢ The color for the column selected is now the purple color. Similar to the data labels, this needs to be
repeated for each column.
➢ Microsoft Excel and Powerpoint usually save the color in the “Recent Colors” section which you will see
when you open the "Colors" dropdown again. You will not have to enter the code each time.

FEMALE TO MALE EMPLOYMENT RATIO FEMALE TO MALE EMPLOYMENT RATIO

0.64 0.64
0.62 0.61 0.61 0.60 0.62 0.61 0.61 0.60
0.53 0.53

Household with children Extended Family

Household with children Extended Family Overall
Overall Couple with children
Couple with children Household without children Couple without children
Household without children Couple without children
 Removing the legend
➢ Now that all columns are the same color, there is no need for a legend. Instead, the labels should be on
the x-axis under each column.
➢ One way to add labels (without significant work in Excel) is to delete the legend and add the labels as
individual text boxes when annotating the visual. This may be preferable if you are already adding the
title and source in another tool (i.e. in MS Word or Powerpoint, Canva/Visme, Adobe Illustrator).
➢ To remove the legend, click on the chart, then on the “Chart Elements” plus symbol. Uncheck “Legend”.
• Before removing the legend, ensure that you have the data value-category combinations saved
somewhere so that you can add the annotations later. You can always refer to the original Excel file.
➢ This visual can then be copied into Word or Powerpoint or any other tool to then add the labels.

FEMALE TO MALE EMPLOYMENT

RATIO
0.64 0.62 0.61 0.61 0.60 0.53

Household with children

Extended Family
Overall
Couple with children
 TASK 5
➢ Go to Sheet named "Task 5" in the Excel file "Training Dataset Day 1".
➢ Adjust the chart in the sheet to change the color of the bars to hex code #1D5934 or RGB code (29, 89, 52).
➢ Adjust the chart in the sheet to change the data label colors so that they are visible against this dark bar
color.
 TRANSFORMATION

BEFORE AFTER

FEMALE TO MALE EMPLOYMENT RATIO

FEMALE TO MALE EMPLOYMENT RATIO
Household with children Extended Family Overall 0.64
0.62 0.61 0.61 0.60
Couple with children Household without children Couple without children
0.53

0.64 0.62 0.61 0.61 0.60

0.53

Household Extended Couple Household Couple

Overall with
Female to Male Employment Ratio with Family without without
children children children children
 What We Do
PRACTICE EXERCISE:
Recreate a visual from trainer’s screen using data from the Training Dataset file
 Exercise – recreate the visual
➢ Using the data in the Sheet named "Exercise 3" in the Excel file "Training Dataset Day 1", recreate the
visual below. The male hex code is #003399 or RGB (0, 51, 153) and the female hex code is #008080 or
RGB (0, 128, 128).

➢ If your visual doesn’t look the same, think of the following:

• Are you using the same chart type?
• Have you adjusted the data structure or created disaggregation categories?
• Have you adjusted the gridlines, data labels, colors, chart title?

➢ Time for exercise: 20 minutes Labor force participation by gender, age, and location
90%
78.4% 78.5% 77.4%
80%
70% 66.5% 66.9%
61.0% 61.2%
60% 53.9% 54.6% 54.5% 55.4%

50%
40% 33.5%
30%
20%
10%
0%
15+ 15-24 15+ 15-24 15+ 15-24
Total Rural Urban

Female Male
What We Do
ANNEX
 Video for widening the plot area
➢ Click the plot area and drag the right side of the area towards the right to make the plot area wider.
• To drag, you must hover over the circular point for the mouse to turn into an arrow (see video).
• You will see a light blue outline if done correctly while you’re dragging the plot area.
• If you happen to move the plot instead of widen it, you can still adjust it – it might take more steps.

WIDEN PLOT
MOVE PLOT
(CORRECT)
 What We Do
TRANSFORMING LABELS
 Moving the data labels to the base of the bars
➢ After removing the legend, instead of adding the category labels in another tool, another option is to
transform the data labels into the category labels.
➢ Click on the chart and then on the “Chart Elements” plus symbol. Then click “Data Labels” and select
“Inside Base”. You should see the data labels move to the bottom. Repeat and select “More Options”.
➢ In the “Format Data Labels” panel on the right side, click the bar chart symbol and then “Label Options”.
1. 3.

2.

4.
 Transforming data labels into category labels
➢ Check the box next to “Series Name” and uncheck the boxes next to “Value” and “Show Leader Lines”
➢ You should now see the name rather than the data label. Repeat for each of the data labels.

3.

1.

2.

4.
 Transforming data labels into category labels
➢ Check the box next to “Series Name” and uncheck the boxes next to “Value” and “Show Leader Lines”.
➢ Then click the “Text Options” tab and change the text color to black.

3.
3.

1.
2.

4.
4.
 Adjusting the size of the plot area
➢ You should now see black category labels instead of a white data labels.
➢ Click the inner plot area with the bars and drag upwards to make the area shorter. Now there is space to
drag down each of the category labels.
 Moving the labels below the plot area
➢ Click each category name separately and drag it down into the white space below the purple columns.
➢ Adjust the sizing of the text by moving the 4 corners of the text box.
➢ Repeat for each category until they are all below the corresponding purple column.

FEMALE TO MALE EMPLOYMENT RATIO

Household Extended Couple Household Couple

Overall with
with Family without without
children children children children
 Changing data labels to category labels
➢ However, readers can no longer tell what values the columns correspond to.
➢ You could either add the Y-axis and/or gridlines back or you could copy the visual (now with category
labels) into Word, Powerpoint, or another tool (Canva, Visme) and add data labels as separate text boxes.

FEMALE TO MALE EMPLOYMENT RATIO FEMALE TO MALE EMPLOYMENT RATIO

70%

60%

50%

40%

30%

20%

10%

0% Couple
Household Extended Household Couple
Extended Household Couple Overall with
Household Couple with with Family without without
Family Overall
with children children without without children children children children
children children
STRENGTHENING
GENDER
STATISTICS

DATA VISUALIZATION
TRAINING MODULE 4:
VISUALIZATION TOOL
COMPARISON
 Data visualization tools comparison
Feature Excel Datawrapper Flourish

Suitable to work with static charts

Suitable to work with interactive charts

Operates in multiple languages

Works without Internet connection

Offers a wide variety of chart options, types and visuals

Offers the possibility to safely encrypt and store data

Has a free version

 Datawrapper
➢ Pros
• Create polished charts and advanced tables in a few
clicks compared to Excel which requires many steps
• Easy to change chart or transpose data structure in
tool without changing original datasheet in Excel.
• Built-in country maps allow easy visualization at
province, administrative region level.
• Easy to annotate, highlight one or a few lines, bars,
data points using specific colors, or to arrange
groups of categories vertically.
➢ Cons
• Chart types are limited compared to Flourish.
• There isn’t always sample data to understand the
data structure for that particular chart type.
• For interactive charts, you must publish the chart
• Internet required.

Training materials available:

https://www.datawrapper.de/training-materials
Datawrapper advanced tables

Training materials available: https://www.datawrapper.de/training-materials

 Flourish
➢ Pros:
• Extensive static and interactive visualization
template library with sample data and visuals for
all charts.
• Easy to create even advanced chart types with
many customization options, and grouping or
layout options.
• Create data stories with multiple types of visuals,
animation, and scrollytelling.
➢ Cons:
• For interactive charts, you must publish the chart.
• Sometimes you must restructure the data to
highlight a certain data point, line, bar, series.
• Compared to Datawrapper, sometimes colors and
annotations are less intuitive – take longer
• Internet required.

Flourish beginner course available:

https://training.flourish.studio/
Flourish grid of charts

Flourish beginner course available: https://training.flourish.studio/

Flourish data stories

Flourish beginner course available: https://training.flourish.studio/

When to use each data tool
Use this tool if/when you…

• do not have access to the internet;

• have data that are not yet public and the data storage/encryption requirements;
• do not suffice for the private/sensitive data.

• want to quickly create a standard static or interactive chart or advanced tables (with
sparklines or data bars);
• prefer to easily highlight one or a few main data points, lines, or bars in the chart;
• want to quickly create a subnational geographic map, range plot, bullet bar;
• need to work from a mobile-phone or tablet.

• need to create a chart type not available in other tools or a more advanced chart type;
• need to see sample data or visuals to understand how to create the visual;
• want to easily create a data story;
• prefer to group data horizontally instead of vertically;
• want to visualize data in “small multiples” or a “grid of charts”;
• prefer to add a highlight line or sort the legend options in a particular order.
 Infographics & social media: Canva and Visme
➢ Canva, Visme, Infogram, Venngage, Piktochart etc. produce a
collection of imagery, text, and data visualizations to give an eye-
catching, creative overview of a topic. All have similar functions.
 Dashboards: Tableau and PowerBI
➢ Tableau and PowerBI provide an interface with interactive visualizations and business intelligence “BI”
capabilities allowing users to create their own dashboards or reports mostly to track/monitor progress.
➢ Dashboards use a combination of interactive visualization types and tables that are sortable, filterable.
➢ Dashboards are exploratory (not explanatory) visualizations – they allow users to explore various
insights rather than highlighting one or w few key insights or narratives/stories.
ANNEX
Comparison of SGS
indicators in each data
visualization tool
Managerial positions
 Time use

Male 4.5 0.6

Female 15.9 3.4

0 5 10 15 20 25

% of day
Unpaid Domestic Work Unpaid Care Work
Employment by sector
Unemployment
Asset Ownership
Ratio of female to male labor force participation
Asset ownership
Gender Pay Gap

Source: Paris21 Communicating Gender Statistics on Women’s Economic Empowerment Course

STRENGTHENING
GENDER
STATISTICS

DATA VISUALIZATION
TRAINING MODULE 5:
EXCEL CHART
TRANSFORMATIONS
 Contents

1. Recap of data visualization steps and breakdown

2. Using data visualization principles to create better charts in Excel:
• Transforming a pie chart into a stacked chart
• Transforming a grouped column chart into a bullet bar or
a line chart
Specific actions for creating better charts
1. Changing chart type of an existing chart (slides 15-16)
2. Adjusting data selection for a chart (slides 17-18)
3. Widening the plot area within a chart (slide 19)
4. Using the “Switch Row/Column” button to reformat a
stacked bar chart (slide 21)
5. Adding a chart title & adjusting the bar thickness (slide 22)
6. Changing the bar colors (slides 24-28)
• Choosing appropriate colors (slide 25)
• Selecting HEX/RGB codes or color palettes (slides 26)
7. Overlapping bars (slides 36-38)
8. Removing secondary axis (slide 41)
9. Adjusting frequency of years on x-axis (slide 48)
10. Adjusting line type (slide 49)
11. Adjusting the size of dots (slide 50)
12. Adjusting the colors of lines and dots (slide 51)
13. Moving the legend and widening the chart (slides 52-53)
1. Recap of data
visualization steps
and breakdown
Recap: steps for creating a data visualization

Explore/try
visualization
options Three-fourths of creating a visualization is add
Optional: cleaning
Design, format, Publish and/or
Upload data and finalizeand formatting the data in the proper additional
download way to:
visualization visualization annotations
as needed
Adjust data match the required data structure, inputs,
structure
and features of the intended visualization type

highlight the right message and insights

enable easy formatting and annotation

Recap: data visualization breakdown

25%
Formatting &
75% annotating
Three-fourths of creating a visualization is cleaning
the visual
Properly and formatting the data in the proper way to:
cleaning,
match the required data structure, inputs,
transposing and features of the intended visualization type
& preparing
the data highlight the right message and insights

enable easy formatting and annotation

Recap: data visualization breakdown
2. Using data visualization
principles in Excel to create
better charts
Enhancing Excel charts
PART 1: ADJUSTING DATA OR PART 2: USING BETTER CHART
CHART ELEMENTS TYPES

Example 1: Example 1:
Grouping and ordering data, chart elements, Transforming pie chart to a stacked column chart.
and legend for the right data presentation.
Example 2:
Example 2: Transforming grouped column chart with multiple
Simplifying (removing unnecessary chart disaggregations into a bullet bar.
elements), intentional and appropriate sorting,
and using the right data visualization attributes Example 3:
to highlight insights (color, shape, pattern, etc.) Transforming grouped bar chart to line chart.

PRACTICE EXCERCISES
1. Recreate a visual from trainer’s screen using data from Training Dataset file.
2. Create one’s own visual with data of choice from Training Dataset file.

NOTE: It is recommended to complete PART 1 and the accompanying Practice Exercise from
Excel Module 3 prior to moving on to PART 2 (this module).
 What We Do
TRANSFORMING A PIE CHART
INTO A STACKED CHART
 GROUP DISCUSSION
➢ Discussion – Can you understand the main insight? What other ways could this visual be shown that
are more efficient? What would you change in this visual?

Reasons for not seeking employment

Female Male

22% 23%
23% 24%

57% 43%
12%

12% 9%
16%
18%

16% 9%
9% 18% 16% 9%

Full time student Too young/Too old

Disabled/Ill My spouse wouldn't allow that

Occupied with home duties Other

 GROUP DISCUSSION
➢ Discussion – Can you understand the main insights better? Which one shows the gender gap better?
Can you still understand the individual reasons as a proportion of all reasons (part-to-whole)?

Reasons for not seeking employment Reasons for not seeking employment
50 50
45
40 40
35
30 30
25
20
20
10 15
10
0 5
Female Male 0
Full time Too young/ Disabled/ Ill My spouse Occupied with Other
Full time student Too young/ too old student too old wouldn't allow home duties
that
Disabled/ Ill My spouse wouldn't allow that
Occupied with home duties Other Female Male
 GROUP DISCUSSION
➢ The 100% stacked bar chart compared to the pie chart…
• shows the values in length which are easier to compare than the pie slices;
• shows the percentage/proportion of the whole;
• is easier to identify gender gap comparison with female visually above male comparing lengths;
• is created in one chart not two separate charts.

Reasons for not seeking employment Reasons for not seeking employment
Female Male

Female
23%
22% 23% 24%

57% 43%
12%

12% 9% Male
16% 18%

16% 9%
9% 18% 16% 9%
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Full time student Too young/Too old

Disabled/Ill My spouse wouldn't allow that
Full time student Too young/Too old
Occupied with home duties Other
Disabled/Ill My spouse wouldn't allow that

Occupied with home duties Other

 Stacked bar and column charts save space
➢ These images are not created in Excel, but serve to explain how pie charts take up a lot of space and can
be converted into stacked charts.
➢ Having two separate pie charts for female and male makes it difficult to easily visualize the gender gap.

Total Rural Urban

 Changing multiple pie charts into stacked chart
➢ Click one of the original pie charts then click the “Insert” tab in the Excel ribbon (top toolbar menu).
➢ Under the "Charts" section click the bar chart looking icon and select the stacked bar chart.

2. 3.

4.
1.
 Changing multiple pie charts into stacked chart
➢ Alternatively, click one of the original pie charts then click the “Chart Design” tab in the toolbar menu at
the top. Then click the “Change chart type” icon and select stacked bar chart.
2. 3.
3.

5.

4.
1.
 Adjusting data selection for a chart
➢ The chart is currently only picking up the "Female" column so it just looks like one bar of data.
➢ To adjust the data range for the chart, right click the chart and select “Select Data”.
➢ The “Select Data Source” box will pop up highlighting the current data range in dashed green lines.

4.
1.
 Readjusting data for stacked bar chart
➢ Highlight the full range of data to include both the female and male columns. Then click “OK” or hit the
"Enter" key.
➢ Now there should be both female and male data in the chart.

Female
Other 22%

Occupied with home duties 16%

My spouse wouldn't allow… 9%

Disabled/Ill 18%

Too young/Too old 12%

Full time student 23%

0% 50% 100%

Male Female
 Widening the plot area
➢ Click the plot area and drag the right side of the area towards the right to make the plot area wider.
• To drag, you must hover over the circular point for the mouse to turn into an arrow (see video).
• You will see a light blue outline if done correctly while you are dragging the plot area.
• If you happen to move the plot instead of widen it, you can still adjust it – it might take more steps.

1. 2.
 TASK 6
➢ Go to Sheet named "Task 6" in the Excel file "Training Dataset Day 2".
➢ Adjust the chart in the sheet by changing one of the two pie charts to a stacked bar chart.
➢ Make sure to readjust the data structure to ensure there are two bar colors on the chart, not one.
 Reformatting the stacked bar chart
➢ The stacked bar chart rows are by “Reason for not seeking employment” when they should be by “Gender”.
➢ To reformat, click the chart, then click the “Chart Design” tab in the Excel ribbon (top toolbar menu).
➢ Then click “Switch Row/Column" and you will see the chart transform.
2.
3.

Female

1.
Female

2. Male

0% 20% 40% 60% 80% 100%

Full time student Too young/Too old

Disabled/Ill My spouse wouldn't allow that
Occupied with home duties Other
 Finalizing the chart title and bar thickness
➢ The chart title needs to be adjusted as well as the color scheme for the stacked bar charts.
➢ Click the chart title and type in the new title.
➢ Click the bar segment and in the “Format Data Series” panel to the right, click the bar icon.
➢ Then under “Series Options” use the slider to adjust the thickness of the bars.
• Especially when there are only two bars, the gap between the bars should be smaller than the bars
themselves. In the image on the left, the gap is 106% and in the image on the right the gap is 57%.

Reasons for not seeking employment by gender

1. 3.
2. Female

4.
Male

0% 20% 40% 60% 80% 100%

Full time student Too young/Too old

Disabled/Ill My spouse wouldn't allow that
Occupied with home duties Other
 TASK 7
➢ Go to the same chart you adjusted in Task 6 in the Sheet named "Task 6" in the Excel file "Training
Dataset Day 2".
➢ Adjust the stacked bar chart by using the "Switch Row/Column" step so that there are only two stacked
bars, one for female and another for male.
➢ Adjust the bar thickness of the stacked bar chart.
 Changing the bar colors
➢ Click the bar segment that corresponds to “Other” which is currently green and change it to grey as
“Other” and “No data” should be neutral colors to minimize attention.
• Click on the paint bucket icon in the “Format Data Series” panel to the right. Under “Fill”, click on the
paint bucket dropdown under “Color” and choose grey.
➢ Repeat the steps to adjust the other bar segment colors. Either select from the colors available or click
“More colors” and add hex codes under the “Custom” tab.

1.
2.

3.
 Choosing appropriate colors
➢ Choose categorical colors (not sequential or diverging).
➢ Choose colors that will work with the rest of the report. Do not reuse colors that have been assigned for
other categories “female”, “male”, “total”, “urban”, “rural”, etc.
➢ For the bar segments that you want to draw attention to like “Occupied with home duties” or “My spouse
wouldn’t allow that” choose darker or brighter colors. Choose lighter colors for those bar segments that
don’t seem to have much of a gender gap like “Too young/too old”, and “Disabled/ill”.

Diverging
Categorical Sequential
 Selecting hex codes and color palettes
➢ Hex codes can be found at htmlcolorcodes.com and https://www.computerhope.com/htmcolor.htm
➢ Color palettes can be found at colorhunt.co or colormind.io
➢ You can also find these by typing searching in an internet search engine like Google Chrome.
➢ The hex code will appear upon hovering over the color or written right underneath the color.
 Choosing appropriate colors
➢ Intentionally choosing colors transforms a visual from looking “outdated” to “polished”.

Reasons for not seeking employment by gender Reasons for not seeking employment by gender

Female Female

Male Male

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
Full time student Too young/Too old Full time student Too young/Too old
Disabled/Ill My spouse wouldn't allow that Disabled/Ill My spouse wouldn't allow that
Occupied with home duties Other Occupied with home duties Other
 Alternative versions
➢ Choose colors that will work with the rest of the report. Do not reuse colors that have been assigned for
other categories “female”, “male”, “total”, “urban”, “rural”, etc.
Reasons for not seeking employment by gender Reasons for not seeking employment by gender

Female Female

Male Male

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
Full time student Too young/Too old Full time student Too young/Too old
Disabled/Ill My spouse wouldn't allow that Disabled/Ill My spouse wouldn't allow that
Occupied with home duties Other Occupied with home duties Other

Reasons for not seeking employment by gender Reasons for not seeking employment by gender

Female Female

Male Male

0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100%
Full time student Too young/Too old Full time student Too young/Too old
Disabled/Ill My spouse wouldn't allow that Disabled/Ill My spouse wouldn't allow that
Occupied with home duties Other Occupied with home duties Other
 TASK 8
➢ Go to the same chart you adjusted in Task 6 and 7 in the Sheet named "Task 6" in the Excel file "Training
Dataset Day 2".
➢ Choose a categorical palette for the stacked bar chart out of the following options.
➢ Use the eyedropper tool to find the hex codes on the color palettes below, then assign a color to each
category.
 TRANSFORMATION

BEFORE AFTER

Reasons for not seeking employment Reasons for not seeking employment by gender

Female Male
Female
22% 23%
23% 24%

57% 43%
12%

12% 9%
16%
18% Male

16% 9%
9% 18% 16% 9%
0% 20% 40% 60% 80% 100%

Full time student Too young/Too old

Full time student Too young/Too old Disabled/Ill My spouse wouldn't allow that
Occupied with home duties Other
Disabled/Ill My spouse wouldn't allow that

Occupied with home duties Other

 What We Do
TRANSFORMING A GROUPED COLUMN CHART
WITH MANY CATEGORIES INTO A BULLET BAR
 GROUP DISCUSSION
➢ Discussion – Can you understand the main insight? What other ways could this visual be shown that
are more efficient? What would you change in this visual based on data visualization principles?

Proportion in tertiary education by gender and field of study

70 63.43 62.00 60 60
57.8 58.2 54.79
60 50.3 49.0846
50
37.5
36.26
40 32.2835
27.00 25.19
30
20
10
0

Female Male
 GROUP DISCUSSION
➢ Discussion – Can you understand the main insights better? Which one shows the gender gap better?
➢ Is there anything you would still change in the visual on the right?

Proportion in tertiary education by gender and field of study Proportion in tertiary education by gender and field of study
70 63.43 62.00
57.8 58.2 60 60 35
60 54.79 Science, Technology , Engineering and Mathematics… 32.28
50.3 49.0846
50 Agriculture, Forestry, Fisheries and Veterinary 46
37.5
36.26 49.08
40 32.2835
27.00 25.19 Engineering, Manufacturing and Construction 60
30 25.19
20 Business, Administration and Law 50.3
54.79
10
Information and Communication Technologies 60
0 27.00
Social Sciences, Journalism and Information 58.2
62.00
Arts and Humanities 57.8
63.43
Natural Sciences, Mathematics and Statistics 37.5
36.26

0 10 20 30 40 50 60 70

Male Female
Female Male
 GROUP DISCUSSION
➢ The bullet bar compared to the clustered column chart:
• is easier to identify the gender gap at a glance with the skinny bar extending past the thicker;
• is cleaner and more consolidated (allows for many disaggregations to fit in one chart);
• allows the category labels to be horizontal for ease of reading;
• brings variety to the visualization types within a report, while still remaining close to the
traditional bar/column chart.

Proportion in tertiary education by gender and field of study (%) Proportion in tertiary education by gender and field of study (%)
70 63.43 62.00 60 60
57.8 58.2 54.79 Arts and Humanities
60 50.3 49.0846
50 Social Sciences, Journalism and Information
36.2637.5
40 32.2835
27.00 25.19 Business, Administration and Law
30
20 Agriculture, Forestry, Fisheries and Veterinary
10
Natural Sciences, Mathematics and Statistics
0
Science, Technology , Engineering and Mathematics

Information and Communication Technologies

Engineering, Manufacturing and Construction

0 10 20 30 40 50 60 70

Female Male
Female Male
 Changing chart type and removing data labels
➢ Click the original chart and then click the “Insert” tab in the Excel ribbon (top toolbar menu).
➢ Under the "Charts" section, click the bar chart looking icon and select the clustered bar chart.
➢ Then click on the chart and the “Chart Elements” plus symbol and uncheck the box next to “Data labels”.
2.
3.

6.
5.
4. 7.

1.
 Overlapping bars to create bullet bars
➢ Overlapping bars will have either the female values on top or the male values or vice versa.
• Whichever bar is on top is thinner and the bar behind it is thicker in order to properly see the gap.
• If the bar behind is thinner, the data may end up being completely hidden by the thicker bar on top.
➢ Click on the bar series that you would like to show on top. The “Format Data Series” panel will show up.
➢ Click on the bar chart icon in order to see the series options.

1.
2.
 Overlapping bars to create bullet bars
➢ After selecting the bars to go on top, under the “Series Options – Plot Series On” switch primary axis to
secondary axis. You will notice the bars are now overlapped and there is another axis on top of the chart.
➢ Under “Gap Width”, move the slider to the right. You will see the orange bars getting thinner. The
percentage in the box next to the slider will increase – you can also enter a percentage in that box (for
example, 430%).

1.

1.

2.
 Overlapping bars to create bullet bars
➢ Select the other colored bar series and go directly to the “Gap Width” section within the “Format Data
Series” panel. Do not touch anything in the “Plot Series On” section.
➢ Under “Gap Width”, this time move the slider to the left. You’ll see the blue bars getting thicker. The
percentage in the box next to the slider will decrease – you can also enter a percentage in that box (for
example, 50%).

2.
1.
 TASK 9
➢ Go to Sheet named "Task 9" in the Excel file "Training Dataset Day 2".
➢ Choose which bar series will go on top.
➢ Adjust the bars so that the the bar series you want on top is overlapping the other bar series.
➢ Adjust the top bar series so they are skinnier than the other bar series.
➢ Adjust the bottom bar series so they are thicker.
 Sorting the bars by value
➢ Click in the column for the variable which you wish to sort. In this case we will sort by female values.
➢ Then under the “Home” tab in the Excel ribbon (top toolbar menu), click on the “Sort & Filter” button in
the “Editing” section. Then click “Sort A to Z” (or “Sort Z to A” for the other direction).
2.
3.

4.
1.
 Finalizing the chart
➢ Click the chart title and add an appropriate chart title.
➢ Remove the secondary axis at the top of the chart by clicking the axis and hitting the delete button.
➢ Adjust the colors as per preference or consistency with the report.

Proportion in tertiary education by gender and field of study (%)

Arts and Humanities

Social Sciences, Journalism and Information

Business, Administration and Law

Agriculture, Forestry, Fisheries and Veterinary

Natural Sciences, Mathematics and Statistics

Science, Technology , Engineering and Mathematics

Information and Communication Technologies

Engineering, Manufacturing and Construction

0 10 20 30 40 50 60 70

Female Male
 TRANSFORMATION

BEFORE AFTER

Proportion in tertiary education by gender and field of study Proportion in tertiary education by gender and field of study (%)
70 63.43 62.00 60 60
57.8 58.2 54.79 Arts and Humanities
60 50.3 49.0846
50 Social Sciences, Journalism and Information
37.5
36.26
40 32.2835
27.00 25.19 Business, Administration and Law
30
20 Agriculture, Forestry, Fisheries and Veterinary
10
0 Natural Sciences, Mathematics and Statistics

Science, Technology , Engineering and Mathematics

Information and Communication Technologies

Engineering, Manufacturing and Construction

0 10 20 30 40 50 60 70
Female Male Female Male
 What We Do
TRANSFORMING A GROUPED COLUMN CHART
INTO A LINE CHART
 GROUP DISCUSSION
➢ Discussion –What other ways could this visual be shown that are more efficient? What do you like
about this visualization? What would you change in this visual based on data visualization principles?
Prevalence of stunting (% of children under 5) by gender and severity
41
37
34 34 35 35
34
30 30 31
29
27

17 18
16 15 15 16
13 13 14 13 14
11

1991 2004 2006 2011 2014 2018 1991 2004 2006 2011 2014 2018
DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI
Stunting Severe stunting

Girls Boys
 GROUP DISCUSSION
➢ Discussion – Can you understand the time series insights better? Is there anything you would still
change in the visual on the right?

Chart Title Chart Title

45
50
40
40 35
30 30
20 25
10 20
15
0
1991 2004 2006 2011 2014 2018 1991 2004 2006 2011 2014 2018 10
DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI 5
0
Stunting Severe stunting
1991 2004 2006 2011 2014 2018
Female Male
Stunting Female Stunting Male
Severe stunting Female Severe stunting Male
 GROUP DISCUSSION
➢ The line chart compared to the side- by-side clustered column chart:
• is easier to identify the widening and narrowing of the gender gap over time;
• looks less busy/cluttered and takes up less space in a report;
• Allows better comparison between the “Stunting” and “Severe stunting” not just the gender gap.

Prevalence of stunting (% of children under 5) by gender and severity

45
Prevalence of stunting (% of children under 5) by gender and severity
40
50
40 35
30
30
20
Stunting Female
10 25
0 Stunting Male
1991 2004 2006 2011 2014 2018 1991 2004 2006 2011 2014 2018 20 Severe Stunting Female
DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI
15 Severe Stunting Male
Stunting Severe stunting
Female Male 10

0
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017
 Changing chart type to line chart
➢ Click the original chart and then click the “Insert” tab in the Excel ribbon (top toolbar menu).
➢ Under the "Charts" section click the line chart looking icon and select the line chart with the dots.
2.
3.

4.

1.
 Adjusting the frequency of years on the x-axis
➢ The current chart is misleading with inconsistent frequency of years along the x-axis. To adjust this, click
the x-axis of the chart with the dates. You should see the “Format Axis” panel on the right side.
➢ In the “Format Axis” panel, click the bar chart icon and open the “Axis Options” section.
➢ Under “Axis Options”, change the setting from “Automatically select based on data” to “Date axis”.
You should now see that the x-axis has equally spaced years from 1991 to 2017.

Chart Title
2. 45
40
35
30
25
3. 20
15
1. 10
5
0
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017

Stunting Female Stunting Male

Severe Stunting Female Severe Stunting Male
 Adjusting the line type
➢ Instead of using four colors, two colors can be used for female/male and two line types can be used for
the other disaggregation severe stunting/stunting.
➢ Click on the line you wish to change and then click the paint bucket icon within the “Format Data Series”
panel on the right. Under "Line" go to the "Dash type" dropdown menu and select your preferred line
type. Here we selected the first dotted line option.
➢ Repeat this for all other lines in the same disaggregation category.

Chart Title
2. 45
1. 40
35
30
25
20
15
10
5
0
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017

3. Stunting Female Stunting Male

Severe Stunting Female Severe Stunting Male
 Adjusting the size of the dots
➢ To avoid misleading readers, it is essential to use dotted line charts to show the gaps in years of data
collection. These dots are called “Markers” and the default size is 5.
➢ To make the dots more visible, click on the line with the dots you wish to change and then click the paint
bucket icon within the “Format Data Series” panel on the right. Click "Marker" and go to "Marker Options".
➢ Change the setting from “Automatic” to “Built-in”. Then under "Size", type in either a number or use the
arrows to increase the size. Here we increased it to 7.
➢ Repeat this until all the dotted lines have the same size “dots”/”markers”.

Chart Title
45
40
35
2. 30
1. 3. 25
20
15
10
5
0
4. 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017

Stunting Female Stunting Male

Severe Stunting Female Severe Stunting Male
 Adjusting colors of the remaining lines and dots
➢ To ensure that the female and male colors are consistent in the chart, the solid lines (i.e. severe stunting)
need to be changed so that "Severe Stunting Female" is blue and the "Severe Stunting Male" is orange.
➢ Click on the yellow line (currently Severe Stunting Male) and then click the paint bucket icon within the
“Format Data Series” panel on the right. Under “Line” go to the color dropdown menu and select orange.
➢ Under “Marker” go to the color dropdown menu under both “Fill” and “Border” and select orange.
➢ Repeat this for the "Severe Stunting Female" grey line, changing both the line and the dots to blue.

2.
3. 5.
1.
4.

6.

7.
 Moving the legend and adding a title
➢ Click the chart title and add an appropriate chart title.
➢ To move the legend to the right side, click on the chart and then on the “Chart Elements” plus symbol.
➢ To the right of “Legend”, click on the arrow. Then select the position of the legend – in this case “Right”.

2.
Prevalence of stunting (% of children under 5) by
1. gender and severity
45
40
35
3. 4. 30
Stunting Female
25
Stunting Male
20
15 Severe Stunting Female

10 Severe Stunting Male

5
0

1991

1995

1999

2003

2007

2011

2015
1993

1997

2001

2005

2009

2013

2017
 Widening the chart
➢ The chart seems a bit squished making the years on the x-axis unreadable at a 90 degree angle.
➢ Pull the left side of the chart at the circle towards the left until you see the years are horizontal again.

Prevalence of stunting (% of children under 5) by gender and

severity
45

40

35
1. 30
Stunting Female
25
Stunting Male
20
Severe Stunting Female
15 Severe Stunting Male
10

0
1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017
 TRANSFORMATION

BEFORE AFTER

Prevalence of stunting (% of children under 5) by gender and severity

41
37
35
34 34
Prevalence of stunting (% of children under 5) by gender and severity 31
34 35
50
40 30 30 Girls Boys
29
30 27
20 18
17
10 16
15 14 Stunting
0 13
1991 2004 2006 2011 2014 2018 1991 2004 2006 2011 2014 2018 16
DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI
15 14 Severe
13 13
11
Stunting
Stunting Severe stunting
Female Male

1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018
 TASK 10
➢ Go to Sheet named "Task 10" in the Excel file "Training Dataset Day 2".
➢ Change the line type for one of the location disaggregation lines from solid to dashed (either for
national or for urban).
➢ Resize the marks (dots) along the line to be larger (size 7 for example).
➢ Change the color of the other disaggregation so that there are only two colors in the chart – one for
female and the other for male.
 What We Do
PRACTICE EXERCISE:
Create a visual of your own with data from the Training Dataset file
 Exercise – create your own visual
➢ Using any data from any sheet in the "Training Dataset Day 1" or "Training Dataset Day 2" Excel files,
create your own visual with more than four data points.
➢ The gender factbook’s colors are
• Female: #006389; RGB (0, 99, 137)
• Male: #00B0AB; RGB (0,176, 171)
➢ Use the "Data Visualization Tip Sheet" handout/Word file to review your visual prior to finalizing.

➢ A few major questions to keep in mind (more details in the "Data Visualization Tip Sheet" handout):
• Is the visual showing the gender-relevant insight or does data need to be transposed or
restructured?

• Does it look clean or cluttered? Are there unnecessary gridlines, labels, axes duplicate encodings?

• Does the visual have proper sorting, grouping, legend order, etc.?

• Do the colors align? Are they intentional?

• Can any aspect of the chart be misleading for the audience?

What We Do
ANNEX
 Video for widening the plot area
➢ Click the plot area and drag the right side of the area towards the right to make the plot area wider.
• To drag, you must hover over the circular point for the mouse to turn into an arrow (see video).
• You will see a light blue outline if done correctly while you’re dragging the plot area.
• If you happen to move the plot instead of widen it, you can still adjust it – it might take more steps.

WIDEN PLOT
MOVE PLOT
(CORRECT)
 What We Do
ADDITIONAL LINE CHART STEPS - LABELS
 Adding data labels
➢ To add data labels, click on the chart and then on the “Chart Elements” plus symbol.
➢ Check the box next to “Data Labels”.

2.

1. 3.
 Changing position of data labels
➢ The data labels are currently overlapping. The position of the data labels should be changed so that the
line above has data labels above the line and the line below has data labels below the line.
➢ Click on the line that needs the data label change and then on the “Chart Elements” plus symbol. Click the
arrow to the right of “Data Labels” and select “Below”. Repeat this for the other blue line.

2.
3.

1.
4.
 Removing gridlines and y-axis labels
➢ Since there are data labels, there is no longer a need for the y-axis labels or gridlines.
➢ To remove the gridlines, click on the chart and then on the “Chart Elements” plus symbol and uncheck
“Gridlines”.
➢ To remove the y-axis labels, click on the y-axis labels and hit the "Delete" button on the keyboard.

2.
1.

3.
STRENGTHENING
GENDER
STATISTICS

DATA VISUALIZATION
TRAINING MODULE 6:
DATAWRAPPER MAPS
Contents
1. Recap of data visualization steps and breakdown
2. Setting up the Datawrapper tool account login
3. Getting started with visualizing maps in Datawrapper
4. Designing, formatting, & annotating the data visualization
5. Downloading the visualization
Specific actions/functions
1. Selecting the chart type (slides 18-21)
2. Uploading data, reuploading data (slides 22-37, 40-42, 82-85)
3. Cleaning and adjusting data structure (slides 28-39)
- in Datawrapper (slides 28 – 33)
- in Excel (slides 34 – 39)
4. Choosing a color scale (slides 45 – 50)
5. Changing the color gradient (slides 51 – 56)
6. Changing the color legend (slides 58-63)
7. Adding legend labels (slides 64-66)
8. Adding data labels (slides 68-71, 87-89)
9. Merging values from two columns into a third column in Excel
using =CONCATENATE() function (slides 72 – 80)
10. Finalizing annotations: title, subtitle, source (slide 89)
11. Downloading the visual (slides 91-95)
1. Recap of data
visualization steps
and breakdown
Recap: steps for creating a data visualization

Explore/try
visualization
options Three-fourths of creating a visualization is add
Optional: cleaning
Design, format, Publish and/or
Upload data and finalizeand formatting the data in the proper additional
download way to:
visualization visualization annotations
as needed
Adjust data match the required data structure, inputs,
structure
and features of the intended visualization type

highlight the right message and insights

enable easy formatting and annotation

Recap: data visualization breakdown

25%
Formatting &
75% annotating
Three-fourths of creating a visualization is cleaning
the visual
Properly and formatting the data in the proper way to:
cleaning,
match the required data structure, inputs,
transposing and features of the intended visualization type
& preparing
the data highlight the right message and insights

enable easy formatting and annotation

 Recap: data structure
➢ The type of visualizations you can create depend on the structure of the data tables (columns and rows).
➢ See below two different structures of the same data points.

A. Grouped by gender B. Grouped by type of work

Gender Unpaid Domestic Work Unpaid Care Work Type of Work Female Male

Unpaid Domestic Work 15.9 4.5

Female 15.9 3.4

Unpaid Care Work 3.4 .6

Male 4.5 .6
 Recap: data structure
➢ The type of visualizations you can create depend on the structure of the data tables (columns and rows)
➢ See below two different structures of the same data points with grouped bar chart examples.

A. Grouped by gender B. Grouped by type of work

Gender Unpaid Domestic Work Unpaid Care Work Type of Work Female Male

Unpaid Domestic Work 15.9 4.5

Female 15.9 3.4

Unpaid Care Work 3.4 .6

Male 4.5 .6

Difference/gap between the types of unpaid work for a given gender. Difference/gap between men’s and women’s time spent on a given
type of unpaid work.
2. Setting up the
Datawrapper tool
account login
 Set up Datawrapper account
➢ Type https://app.datawrapper.de/ in your internet browser.
➢ Click "Create a new account" in the bottom of the grey box.
➢ Sign up with your email or with another account like Google, Microsoft, Github, or Twitter.
 Set up Datawrapper account
➢ Enter your email address and create a password at least 8 characters long.
 Set up Datawrapper account
➢ Once you have created an account, it will automatically log you in. You should see this screen.
➢ In the right hand corner you will see a message alerting you that your email address needs to be
confirmed.
➢ Check your inbox for the Datawrapper email.
 Confirm Datawrapper account
➢ Once you find the email from Datatwrapper in your inbox, open it. It should look like this.
➢ Click the "Confirm your account" link in the email.
➢ Then log in to your Datawrapper account again.
 Datawrapper account successfully confirmed
➢ After logging in you’ll see a green box in the top right that confirms you have successfully activated
your email address for Datawrapper.
3. Getting started
with visualizing maps
in Datawrapper
 Log in to Datawrapper
➢ Type https://app.datawrapper.de/ in your internet browser.
➢ Click "Sign in with Email".
 Log in to Datawrapper
➢ Enter your email address and password.
➢ Click the blue "Login" button.
 Select a chart category
➢ Click the "Create New" button or "Nouveau" button.
➢ Then select the chart type: Chart, Map, or Table. For this first visual we will select "Map".
 Select a map type
➢ There are three map types: Chloropleth map, Symbol map, and Locator map.
➢ Select the "Chloropleth map".
 Select a geographic region
➢ In the search bar, type a geographic region (i.e. world, region name, country name).
➢ For our visual, we will type in Cameroon to find the Cameroonian maps.
 Select the geographic map level for Cameroon
➢ There are two geographic map levels: Departments and Provinces.
➢ Choose the level for which you have data. We will select "Provinces". You will see a preview of the map.
➢ Then click the "Proceed" button.
 What We Do

UPLOADING DATA
 Upload data from Excel or CSV file
➢ Upload an Excel or CSV file by clicking the "Upload file" button. The box to select the file will pop up.
➢ Select the file from the folder on your computer and click "Open".
 Upload data from Excel or CSV file
➢ A successful upload will show a green check mark next to the "Upload file" button.
➢ If there is more than one sheet in the Excel file, there will be a dropdown to select which sheet to use.
 Match the data with the map
➢ Next to the "Upload" tab, click the "Match" tab.
➢ Make sure that the matching key is set to the right category (i.e. Name vs. Postal code for the province).
➢ Select and/or verify the column for the "Name" as well as the column for the "Values".
 Check the data
➢ Next to the Match tab, click the Check tab. You’ll notice there is a yellow warning symbol and a yellow
warning box that shows two geographic areas do not match with data.
➢ In red, 4 region names do not match: Centre (Sands Yaounde), Yaounde, Littoral (Sans Douala), Douala.
➢ The region names need to be adjusted in the data table to match.
 Required map data structure
➢ The current data structure separates data for cities Douala and Yaounde from their provinces.
➢ The map only takes data for provinces so the data structure must be adjusted in Datawrapper or Excel.

A. Current Data Structure B. Required Data Structure for Map

Birth Birth
Region Region
Registration Registration
Adamaoua 38,7 Adamaoua 38,7

Centre (Sans Yaoundé) 42,5 Centre 42,5

Douala 78,9 Est 32,3

Est 32,3 Extrême-Nord 35,8
Extrême-Nord 35,8
Littoral 54,1
Littoral (Sans Douala) 54,1
Nord 31,5
Nord 31,5 Nord-Ouest 53,9
Nord-Ouest 53,9 Ouest 71,2
Ouest 71,2 Sud 50,7
Sud 50,7 Sud-Ouest 72,8
Sud-Ouest 72,8
Yaoundé 70,1
 What We Do
CLEANING/ADJUSTING THE DATA
OPTION 1: DATAWRAPPER
 Filling in blank cells of data in Datawrapper
➢ Copy or type the values for Centre (Sans Yaounde) and Littoral (Sans Douala) into the blank cells in the
value column for the regions Centre and Littoral respectively.
 Filling in blank cells of data in Datawrapper
➢ When the values area added in the blank cells, the rows will be numbered.
➢ The yellow warning box will disappear as well as the phrase “2 unused” at the bottom of the sheet.
 Removing rows of data in Datawrapper
➢ The rows of data that are no longer needed and still appear red should be deleted.
➢ Highlight the entire row by clicking the row number on the left. It should highlight the row in blue.
➢ It should also show a red box with a trash can symbol. Click on the symbol to delete the row.
 Removing rows of data in Datawrapper
➢ You should then see that the row has disappeared and a popup signifying that you deleted a row.
• In case you deleted the wrong row, you can retrieve it by clicking the “Revert” button.
➢ One of the red boxes (corresponding to the deleted region row) should also have disappeared and the
phrase at the bottom of the sheet should have changed from “4 errors” to “3 errors”.
 Removing rows of data in Datawrapper
➢ Repeat for all rows of data that must be deleted until there are no red “errors” at the bottom of the
datasheet and the red boxes to the left have become a message in green that “All map regions were
matched to a row in your dataset”.
➢ Click the "Proceed" button.
 What We Do
CLEANING/ADJUSTING THE DATA
OPTION 2: EXCEL
 Recap: data structure required by map
➢ The current data structure separates data for cities Douala and Yaounde from their provinces.
➢ The map only takes data for provinces so the data structure must be adjusted in Datawrapper or Excel.

A. Current Data Structure B. Required Data Structure for Map

Birth Birth
Region Region
Registration Registration
Adamaoua 38.7 Adamaoua 38.7

Centre (Sans Yaoundé) 42.5 Centre 42.5

Douala 78.9 Est 32.3

Est 32.3 Extrême-Nord 35.8
Extrême-Nord 35.8
Littoral 54.1
Littoral (Sans Douala) 54.1
Nord 31.5
Nord 31.5 Nord-Ouest 53.9
Nord-Ouest 53.9 Ouest 71.2
Ouest 71.2 Sud 50.7
Sud 50.7 Sud-Ouest 72.8
Sud-Ouest 72.8
Yaoundé 70.1
 Adjust the region names (Edit cell text in Excel)
➢ For the region « Centre (Sans Yaounde) », adjust the name so it says only « Centre ».
➢ For the region « Littoral (Sans Douala) », adjust the name so it says only « Littoral ».
➢ This can be done by clicking into the cell and deleting everything in the parentheses.
 Remove unnecessary rows in Excel
➢ The cities Douala and Yaounde still show up as their own rows of data. They must be deleted.
➢ To delete a row, highlight the entire row by clicking the row number on the left. It should highlight the
entire row in a shaded grey.
➢ Right click on the highlighted row and select “Delete”.

1.

2.
 Remove unnecessary rows in Excel
➢ You should then see that the row has disappeared.
• In case you deleted the wrong row, you can “undo” the action by clicking on the “Undo” button in
the toolbar menu at the top.
➢ The final dataset after removing both rows for Douala and Yaounde is on the right side.
 Save the data in Excel
➢ Save the Excel file so that the changes are not lost.
• It can be saved with the save symbol or by clicking File in the toolbar menu on the top.
• If you do not want to overwrite the Excel file, click Save As instead of Save and save the file under a
different name in your preferred location. For example, Cameroon Map Data Updated. Click "Save".
 Reupload data to Datawrapper
➢ Option 1: Reupload Excel/CSV file by clicking the "Upload file" button, selecting a file, and clicking "Open".
➢ Option 2: Copy the data from the Excel file and paste it in the blank box and click the arrow on the right.

Option 1

Option 2
 Successful upload of the correct data structure
➢ With the Copy/Paste method, there will be a popup notifying that data was pasted with option to revert.
➢ To verify that all data are correctly uploaded or pasted, review both the "Match" and "Check" tabs again
and adjust accordingly if needed.
 Successful upload of the correct data structure
➢ The "Check" tab shows a green check mark and a green box that says "All map regions were matched to a
row in your dataset"
➢ Click the "Proceed" button.
4. Designing and
formatting the visual
 Default map visual
➢ In the "Visualize" tab, you will already see a preview of your map with the default elements.
➢ On the left are buttons to customize the colors, legend, and layout.
 What We Do

CHOOSING A COLOR SCALE

 The default color scale
➢ The default is a linear continuous color scale with a green/blue color palette ranging from the minimum
value of the dataset to the maximum value of the dataset.
➢ Values are automatically assigned a color based on the range. If the range is changed to the minimum and
maximum possible for percentages (0 and 100), then variations will not be visible (see right map).

Min: 31.5 Max 72.8 Min: 0 Max: 100

 Continuous vs. Step scale
➢ To have more control over the color scale, under “Colors” within the “Type” section, change “continuous”
to “steps”. This will produce steps or buckets of colors that can be adjusted by you.
➢ You can choose the number of buckets, the way values are assigned per buckets, or completely customize
the bucket ranges.

Continuous: assigns each value Steps: assigns all values within a specified
a different color on the gradient range (bucket/step) the same color
 Default step color scale: linear
➢ Under “Colors” within the “Steps” section, the default number of steps is 5 and can be changed up to 20.
➢ The default steps are “linear” meaning each range is equidistant between the minimum and maximum.

Ranges Range size

31.5-39.76 8.26
39.76-48.02 8.26
48.02-56.28 8.26
56.28-64.54 8.26
64.54-72.8 8.26
 Adjusting color steps: quantiles
➢ The “linear steps” can be changed to “quantiles”, meaning there is an equal number of observations in
each step.
➢ In the quantile range below, for example, there are two regions per step/color.

5 Quantile Ranges Range size

31.5-35.1 3.6
35.1-40.98 5.88
40.98-51.98 11
51.98-57.52 5.54
64.54-72.8 15.28
 Adjusting color steps: custom
➢ You can also customize the ranges and set the beginning and end value of each range.
➢ Under "Colors", within the “Steps” section select “Custom” from the dropdown list. Then enter the custom
beginning and end values for each range.
 What We Do

CHANGING THE COLOR GRADIENT

 Selecting a different color palette
➢ Under "Colors", within the “Select palette” section, click the color dropdown list.
➢ Select a different color gradient. Remember some are sequential and others are diverging.

Sequential

Diverging
 Selecting a different color palette
➢ Under "Colors", within the “Select palette” section, click the color dropdown list.
➢ Select a different color gradient. Remember some are sequential and others are diverging.

Sequential

Diverging
 Customizing the color palette
➢ If you don’t like the color palette options or must match the branding for a report, they can be changed.
➢ Under Color, within the “Select palette” section, click the wrench button.
➢ A color gradient with several color stops will pop up. Click the reverse button to draw attention to low
values rather than high values.
 Customizing the color palette
➢ Change the color gradient by sliding one or more of the color stops left or right.
➢ Here by moving the color stops towards the green (left) section, the whole color gradient shifts towards
the blue hues than the green ones.
 Customizing the color palette
➢ Change the color gradient by clicking one or more color stops and selecting new colors or entering the
hex codes of the desired color.
➢ On the left, 4 color stops from the original green/blue color palette were changed so that they are all
shades of blue (no green).
➢ On the right, all the color stops were changed to different shades of red.

Adjusting
the color

HEX code
 TASK
➢ In the "Refine" tab, choose a color gradient and change the color scale.
➢ The version you submit should have different colors and scale than what is used in the demonstration
walk-through.
 What We Do

CHANGING THE COLOR LEGEND

 Moving the color legend (steps scale)
➢ Under “Legend” in the “Position” section select a different position from the dropdown.
➢ Steps example:
 Making the color legend vertical (steps scale)
➢ Under “Legend” in the “Orientation” section switch from horizontal to vertical.
➢ The steps are now neatly vertically ordered from lowest values to highest values.
 Moving the color legend (continuous scale)
➢ Under “Legend” in the “Position” section select a different position from the dropdown.
➢ Continuous example:
 Making the color legend vertical (continuous)
➢ For continuous scale in this case moving the legend to the top right is too close to the map so it might be
best to make it vertical.
 Adjusting the size of the color legend
➢ Under “Legend” within the “size” section, you can change the size by sliding the bar or entering a number.
➢ The legend can also be moved slightly from its position under “Legend within the “Offset” section by
typing in a number up to 50 in the horizontal box (which moves it inward). Numbers in the vertical box
would move the legend up or down from its current position.
 What We Do

ADDING LEGEND LABELS

 Adding legend labels
➢ Under “Legend” in the “Labels” switch from “range” to “custom”.
➢ It will show a box for Min and Max and potentially for Center if you already specified a median number.
➢ The default text will say “Low”, “Medium”, “High”
 Adding legend labels (gender gap)
➢ Generally, for gender gaps the values range from negative to positive, which may be confusing to readers.
➢ When charting gender gaps, you always want to change the legend labels/caption to be more user-friendly.
➢ Gender gap examples below with diverging color scale vertical position so labels don’t overlap:
 TASK
➢ In the "Refine" tab, adjust the legend so that it is not using the default position or orientation. Move and
change the legend, position, orientation, and (if you prefer) the labels as well.
 What We Do

ADDING DATA LABELS

 Adding data labels
➢ To add data labels, switch from the “Refine” tab to the “Annotate” tab.
➢ Under the “Map Labels” section, click “Show labels”
➢ The default is to add names of places, like cities.
 Adding data labels
➢ Under “Map Labels”, within the “Type” section, switch from “places” to “columns” which will allow you to
select from the dropdown which column within the dataset to display values for.
➢ The column named Birth Registration shows the data values. Region would show only region names.
 Adding data labels (with regions and values)
➢ To add labels with both region names and data values, there needs to be another column in the dataset
with this combined information (similar to the previous data label exercise in Excel).
➢ This column can be created in Excel and then the data can be reuploaded into Datawrapper as before.

B. Required Data Structure for Map

A. Required Data Structure for Map
with Region and Data Labels
Birth Region Birth Registration
Region
Registration
Adamaoua 38.7 Adamaoua 38.7

Centre 42.5 Centre 42.5

Est 32.3 Est 32.3

Extrême-Nord 35.8 Extrême-Nord 35.8

Littoral 54.1 Littoral 54.1

Nord 31.5 Nord 31.5

Nord-Ouest 53.9 Nord-Ouest 53.9
Ouest 71.2 Ouest 71.2
Sud 50.7 Sud 50.7
Sud-Ouest 72.8 Sud-Ouest 72.8
 WhatFROM
MERGING VALUES We Do
TWO COLUMNS
INTO A NEW COLUMN IN EXCEL:
COMBINED REGION AND DATA LABELS
 Adding a third column title
➢ Open the Excel file and go to the sheet with the data for the correct regions for this map.
➢ Click into the top cell of the column (C) to the right of the “Birth registration” column (B) with data values.
➢ Enter the name of the new column. It can be “Data labels” to easily identify the column later.
 Merging information with =CONCATENATE()
➢ Instead of using a copy/paste method to copy information from column A and B into column C, there is a
function CONCATENATE(cell1,cell2) that can be entered into the cells for column C.
➢ Double click into the cell in the second row and enter the following: =CONCATENATE(
➢ After typing the parenthesis (, click directly into the cell in the second row of the first column. It should
highlight the cell and add the cell number into the function in the third column. =CONCATENATE(A2
• It should look like the image on the right. Do NOT click anywhere else.
 Using the =CONCATENATE() function
➢ After successfully adding the first cell (A2) in the previous step, then type: ,” “,
• The function should look like =CONCATENATE(A2,“ “,
▪ This will add a space between the information in column A and column B.
▪ It should not look like =CONCATENATE(A2,””, with the quotation marks next to each other or else
it will not provide the space between the region name and birth registration value.
➢ It should look like the image on the right. Do NOT click anywhere else.
 Using the =CONCATENATE() function
➢ After successfully adding the quotations,(,” “,) in the previous step, then click directly into the cell in the
second row of the second column.
• It should highlight the cell and add the cell number into the function in the third column. The
function should look like =CONCATENATE(A2,“ “,B2
➢ It should look like the image on the right. Do NOT click anywhere else.
 Result of the =CONCATENATE() function
➢ After successfully adding the cell of the second row in the second column(B2) in the previous step, then
type a closing parenthesis ) and hit enter.
• It should look like this before you hit enter =CONCATENATE(A2,“ “,B2)
➢ The final after hitting enter should be the value from the second row in the first column Adamaoua and the
value from the second row of the second column 38.7 separated by a space: Adamaoua 38.7
➢ It should look like the image on the right.
Full video of =CONCATENATE() function
 Replicating the function for multiple rows
➢ After successfully merging or concatenating information into the first row of the third column, repeat this
for all rows in the column.
➢ The easiest way to repeat is to drag the function down for the whole column.
➢ Click (do not double click) on the cell with the result/function (second row of the third column).
➢ Then in the bottom right corner of the same cell, click and drag until the last row of data. You should see a
green outline when you’re still clicking, but when you release the mouse, you’ll see the results in each row.
 Video replicating the function for multiple rows
➢ After successfully merging or concatenating information into the first row of the third column, repeat this
for all rows in the column.
➢ The easiest way to repeat is to drag the function down for the whole column.
➢ Click (do not double click) on the cell with the result/function (second row of the third column).
➢ Then in the bottom right corner of the same cell, click and drag until the last row of data. You should see a
green outline when you’re still clicking, but when you release the mouse, you’ll see the results in each row.
 TASK
➢ Go to Sheet named "Exercise 5 DW Map" from the "Training Dataset Day 2" Excel file.
➢ Merge values from the first two columns into the third column using the =CONCATENATE() function.
 What We Do

REUPLOADING DATASET WITH DATA LABELS

 Reupload data to Datawrapper
➢ Go to the « Add your data » tab.
➢ Option 1: Reupload Excel/CSV file by clicking the Upload file button, selecting a file, and clicking Open.
➢ Option 2: Copy the data from the Excel file and paste it in the blank box and click the arrow on the right.

Option 1

Option 2
 Successful upload of the revised dataset
➢ With the Copy/Paste method, there will be a popup notifying that data was pasted with option to revert.
➢ To verify that all data are correctly uploaded or pasted, review both the Match and Check tabs again and
adjust accordingly if needed.
 Successful upload of the revised dataset
➢ The Check tab shows a green check mark and a green box that says « All map regions were matched to a
row in your dataset »
➢ Click the proceed button.
 What We Do
FINALIZING ANNOTATIONS
(TITLE, SOURCE, NEW DATA LABELS)
 Return to the annotation tab of visualization
➢ Go to the “Visualize tab” and then switch from the “Refine” tab to the “Annotate” tab.
➢ Under “Map Labels” within the “Select column” section, you will see the “Birth Registration” column is
still selected.
1.

2.
 Select new column for data labels
➢ Under “Map Labels” within the “Select column” section, select the name of the column with the
combined region and data labels you just created.
➢ In the example case, this column is named “Data labels”. The new labels should appear on the map.
 Adding the title, description, and source
➢ Under the “Title” section, add a title in the blank box provided.
➢ Under the “Description” section, add any additional details or subtitles in the box provided.
➢ Under the “Data source” section”, add the name and year of survey in the box provided.
5. Downloading the visual
 Download the visual
➢ Go to the "Publish & Embed" section after finalizing your visual.
➢ Click the PNG image to view the download options. 1.

2.
 Download the visual
➢ Under “Include”, choose whether to download “Just chart” or the chart with “Full header and footer”.
➢ Under “Background”, choose whether you would like the background to be included (white) or
transparent (no background).
➢ Then click “Download image”.
 Download option examples: full header/footer
➢ “Transparent” background is a good option for adding the visual to a social media card or infographic that
may have a colored background.
 Download option examples: just chart
➢ “Just chart” is the best option if you want to add your own annotations or if the title and source are being
added directly in the report and not in each chart.
 A note on publishing
➢ The visual does NOT need to be published in order to download a static image file.
➢ If you wish to create an interactive online version with a URL, you will first need to click Publish Now.
➢ When you publish a visual, the data will be visible to those who have the URL. Data cannot be
downloaded, but the data values can be seen on the data labels and the tooltip if either are enabled.
 What We Do
PRACTICE EXERCISE 5:
Create a map in Datawrapper with data from the Training Dataset file
 Exercise – create a map in Datawrapper
➢ Using the data in Sheet "Exercise 5 DW Map" from "Training Dataset Day 2" Excel file, create a map in
Datawrapper.

➢ Tips to remember:
• The Gender Parity Index indicator has values above and below 1 with 1 indicating equality (gender
parity).
• Diverging colors scales are used for diverging number scales.
• Captions, annotations, notes, etc. should be used if the data must be interpreted in a certain way
by the reader that may not be obvious.

➢ Time for exercise: 20 minutes.

STRENGTHENING
GENDER
STATISTICS

DATA VISUALIZATION
TRAINING MODULE 7:
DATAWRAPPER
RANGE PLOTS
Contents
1. Recap of data visualization steps and breakdown
2. Setting up the Datawrapper tool account login
3. Getting started with visualizing maps in Datawrapper
4. Designing, formatting, & annotating the data visualization
5. Downloading the visualization
Specific actions/functions
1. Selecting the chart type (slides 18-19)
2. Exploring a sample dataset (slides 20-22)
3. Uploading data (slides 23-25)
4. Changing the chart type (slides 27 – 34)
5. Sorting data values (slides 35-37)
6. Finalizing annotations: title, subtitle, source (slides 38-39)
7. Downloading the visual (slide 40-43)
1. Recap of data
visualization steps
and breakdown
Recap: steps for creating a data visualization

Explore/try
visualization
options Three-fourths of creating a visualization is add
Optional: cleaning
Design, format, Publish and/or
Upload data and finalizeand formatting the data in the proper additional
download way to:
visualization visualization annotations
as needed
Adjust data match the required data structure, inputs,
structure
and features of the intended visualization type

highlight the right message and insights

enable easy formatting and annotation

Recap: data visualization breakdown

25%
Formatting &
75% annotating
Three-fourths of creating a visualization is cleaning
the visual
Properly and formatting the data in the proper way to:
cleaning,
match the required data structure, inputs,
transposing and features of the intended visualization type
& preparing
the data highlight the right message and insights

enable easy formatting and annotation

 Recap: data structure
➢ The type of visualizations you can create depend on the structure of the data tables (columns and rows)
➢ See below two different structures of the same data points.

A. Grouped by gender B. Grouped by type of work

Gender Unpaid Domestic Work Unpaid Care Work Type of Work Female Male

Unpaid Domestic Work 15.9 4.5

Female 15.9 3.4

Unpaid Care Work 3.4 .6

Male 4.5 .6
 Recap: data structure
➢ The type of visualizations you can create depend on the structure of the data tables (columns and rows)
➢ See below two different structures of the same data points with grouped bar chart examples

A. Grouped by gender B. Grouped by type of work

Gender Unpaid Domestic Work Unpaid Care Work Type of Work Female Male

Unpaid Domestic Work 15.9 4.5

Female 15.9 3.4

Unpaid Care Work 3.4 .6

Male 4.5 .6

Difference/gap between the types of unpaid work for a given gender. Difference/gap between men’s and women’s time spent on a given
type of unpaid work.
2. Setting up the
Datawrapper tool
account login
 Set up Datawrapper account
➢ Type https://app.datawrapper.de/ in your internet browser.
➢ Click "Create a new account" in the bottom of the grey box.
➢ Sign up with your email or with another account like Google, Microsoft, Github, or Twitter.
 Set up Datawrapper account
➢ Enter your email address and create a password at least 8 characters long.
 Set up Datawrapper account
➢ Once you have created an account, it will automatically log you in. You should see this screen.
➢ In the right hand corner you will see a message alerting you that your email address needs to be
confirmed.
➢ Check your inbox for the Datawrapper email.
 Confirm Datawrapper account
➢ Once you find the email from Datatwrapper in your inbox, open it. It should look like this.
➢ Click the "Confirm your account" link in the email.
➢ Then log in to your Datawrapper account again.
 Datawrapper account successfully confirmed
➢ After logging in you’ll see a green box in the top right that confirms you have successfully activated
your email address for Datawrapper.
3. Getting started
with visualizing maps
in Datawrapper
 Log in to Datawrapper
➢ Type https://app.datawrapper.de/ in your internet browser.
➢ Click "Sign in with Email".
 Log in to Datawrapper
➢ Enter your email address and password.
➢ Click the blue "Login" button.
 Select a chart category
➢ Click the "Create New" button.
➢ Then select the chart type "Chart".
 See an example of the data structure for a range plot
➢ There is an example dataset in Datawrapper for range plots.
➢ Click "Select a sample dataset" in the dropdown menu and select the dataset "Gender Pay Gap"
under Range Plot.
 See an example of the data structure for a range plot
➢ You will see that there are salaries for men and women by level of education for the United States
in this sample dataset.
 See an example of the data structure for a range plot
➢ Go to the “Visualize” section to see what the sample dataset looks like.
 See an example of the data structure for a range plot
➢ Return to the “Check & Describe” section to see how the dataset should look when you upload it
for the gender pay gap data.
➢ Based on this you can now create your own dataset in this format and go back to the “Upload
Data” section and paste the values into the white box or upload an Excel or CSV.
 What We Do

UPLOADING DATA
 Uploading data
➢ Option 1: Copy the data from the Excel file and paste it into the empty box. Click the “Proceed”
button in the bottom right.
 Uploading data
➢ Option 2: Upload your Excel/CSV file by clicking on the "XLS/CSV Upload" button then "Upload a
file", selecting a file, and then clicking "Open".
4. Designing and
formatting the visual
 What We Do

CHANGING THE CHART TYPE

 Changing the graph type
➢ Go to the “Visualize" section. This is the default chart type chosen for the data.
➢ This type of chart is not appropriate for the data. Select a different chart type by clicking on one
of the chart images.
 Changing the graph type
➢ Select the “Range Plot" chart.
➢ Now you'll see that the data are visualized in a range plot, but it's not in order.
 What We Do

CHANGING THE VISUALIZATION ELEMENTS

 Changing the visualization elements
➢ Select the "Refine" tab to display options for editing visualization elements.
➢ Make sure that in the “Range" section, the drop-down menu for “Range start" and the drop-down
menu for “Range end" are not the same. Depending on your preference, you can choose
“Female" for "Start" and “Male" for "End" or vice versa.
 Changing the visualization elements
➢ Under “Labels”, you can choose which data labels to display by toggling on the “Show Values”
button and then making a choice under “Visibility”.
➢ Usually, you want to select to display “both”, which has the data labels for the start and end values.
➢ Sometimes this makes the graph too cluttered, so you can choose “difference” to show the gender
gap.
 Changing the visualization elements
➢ Sometimes selecting “both” makes the graph too cluttered. Here are the options for “difference”
and “% change”.
Difference
Both

% change
 Changing the visualization elements
➢ Under "Labels," click "Label First Range."
➢ You can now see that the category labels for “Female” and “Male” are visible.
 What We Do

SORTING THE VALUES IN THE CHART

 Sorting the values in the chart
➢ Under the "Sorting &
Grouping" section, toggle on
the "Sort Rows” button.
➢ There are four options for
sorting values. "Start", "End",
"Difference", and "% change".
Usually, the chart is sorted by
"start" or "end," which means
“Female" or “Male."
➢ This helps readers visually see
the order.
 Sorting the values in the chart
➢ You can reverse the order of
the values by toggling on the
"Reverse order” button.
➢ It's best to show this way -
the higher values above and
the lower values are lower.
 What We Do

FINALIZING THE CHART

 Finalizing the chart
➢ Select the "Annotate" tab to finalize the chart with annotations.
➢ In addition to a title, it's important to add a subheading to explain the values. In this case, it's local
currency.
➢ Under "Title," add a title in the empty box provided.
➢ Under "Description," add additional details or subheadings in the box provided.
➢ Under "Data source," add the name and year of the survey in the box provided.
5. Downloading the visual
 Downloading the visualization
➢ The download options are the same
as for the map exercise.
➢ Go to the "Publish & Embed" tab after
finalizing your design.
➢ Click the PNG image to view the
download options.
➢ Under "Include," choose to download
"just chart" or the chart with "full
header and footer."
➢ Under "Background," choose whether
you want the background to be
included (white) or transparent (no
background).
➢ Then click on “Download image".
 Download option examples: full header/footer
➢ “Transparent” background is a good option for adding the visual to a social media card or infographic that
may have a colored background.
 Download option examples: just chart
➢ “Just chart” is the best option if you want to add your own annotations or if the title and source are being
added directly in the report and not in each chart.
 A note on publishing
➢ The visual does NOT need to be published in order to download a static image file.
➢ If you wish to create an interactive online version with a URL, you will first need to click Publish Now.
➢ When you publish a visual, the data will be visible to those who have the URL. Data cannot be
downloaded, but the data values can be seen on the data labels and the tooltip if either are enabled.
 What We Do
PRACTICE EXERCISE 6:
Create a range plot in Datawrapper with data from the Training Dataset file
 Exercise – create a range plot in Datawrapper

➢ Using the data from the "Exercise 6 DW Range Plot" worksheet in the "Training Dataset Day 2" file,
create a range plot in Datawrapper.

➢ Tips to remember:
➢ Select the data labels.
➢ Make sure the category labels are visible.
➢ Make sure the chart is sorted correctly.
➢ Change the colors to the appropriate female and male colors. Don't use stereotypical colors.

➢ Time for exercise: 20 minutes.

STRENGTHENING
GENDER
STATISTICS

DATA VISUALIZATION
TRAINING MODULE 8:
ANNOTATIONS
Contents
1. Recap of data visualization steps and breakdown
2. Using annotations to highlight key information
a. Adding text boxes
b. Incorporating icons
c. Adding lines or shaded shapes to highlight information
3. Exercise to annotate visuals created in Day 1 and Day 2.
1. Recap of data
visualization steps
and breakdown
Recap: steps for creating a data visualization

Explore/try
visualization
options Three-fourths of creating a visualization is add
Optional: cleaning
Design, format, Publish and/or
Upload data and finalizeand formatting the data in the proper additional
download way to:
visualization visualization annotations
as needed
Adjust data match the required data structure, inputs,
structure
and features of the intended visualization type

highlight the right message and insights

enable easy formatting and annotation

Recap: data visualization breakdown

25%
Formatting &
75% annotating
Three-fourths of creating a visualization is cleaning
the visual
Properly and formatting the data in the proper way to:
cleaning,
match the required data structure, inputs,
transposing and features of the intended visualization type
& preparing
the data highlight the right message and insights

enable easy formatting and annotation

Recap: data visualization breakdown
2. Using annotations to
highlight key insights
 What We Do

ADDING LABELS OR TEXT

 Adding annotations through text boxes
➢ Under the "Insert" tab in Excel or PowerPoint, click on "Text Box" within the "Text" section and position
the cursor or text box where you would like your annotation. Type the information (data point value,
category label, etc.) Repeat for each bar/label.
1.
2.
 TASK
➢ Go to Sheet named "Demo 9 Text Boxes" with the bar chart we created from Module 3.
➢ Add data labels and/or category labels to the chart.
• Add them directly in Excel or copy and paste the chart to PowerPoint and add the annotations in
PowerPoint.

FEMALE TO MALE EMPLOYMENT RATIO

Extended Household Couple

Household Overall Couple with
Family without without
with children children
children children
 What We Do

INCORPORATING ICONS
 Using icons for highlighting statistics
➢ Icons can be used to create pictograms using the same repeated icon and coloring/shading to show a
proportion like 2 in 5 women representing 40% or a ratio "For every one man there are 5 women who…".

20%
 Using icons for highlighting statistics
➢ Infographic-like visuals with just images and numbers can be developed for incorporation in a report, an
executive summary, a social media card, etc.
 Using icons for highlighting statistics
➢ Icons can enhance simple visuals like bar charts or pie charts that have very few data points (two or three
data points).
 Using icons for highlighting statistics
➢ Icons can enhance simple visuals like bar charts or pie charts that have very few data points (two or three
data points).
 Using icons for highlighting statistics
➢ Icons can enhance simple visuals like bar charts or pie charts that have very few data points (two or three
data points).
 Using icons for legends
➢ Icons can be added to color legends instead of just having colored or patterned boxes for the legend.

45
40
35
30
25
20
15
10
5
0
1991 2004 2006 2011 2014 2018 1991 2004 2006 2011 2014 2018
DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI DHS-I DHS-II DHS-III DHS-IV DHS-V DHS-VI
Stunting Severe stunting

Girls Boys
 Inserting an icon
➢ To add an icon to a chart, go to the “Insert” tab in Excel or PowerPoint, click on “Icons” within the
“Illustration” section. A box will pop up of icons to choose from and you can enter a key word or choose a
topic to search for an appropriate icon. Once you have found an icon, click the “Insert” button.
1.
2.
 Searching for and choosing an icon
➢ On the left I have selected the “Landscape” topic button. On the right I typed in “Rural” in the search bar
• Sometimes you have to use synonyms to find more options perhaps “Farm”, “Land”, “Agriculture”,
etc. Or instead of typing “Urban” typing “Building”, “City”, “Road”, “Street”, etc.
➢ There are usually two versions of the same icon – one outlined and one filled.
• The choice should depend on whether the icon is visible enough with just the outline, or it needs to
be more prominent.
 Searching for and choosing an icon
➢ When creating a pictogram, typically both outlined and filled icons are selected to demonstrate the
proportion out of the whole.
• For example, 2 filled and 3 outlined versions of the same icon should signify 2 out of 5 or 40%.
• Alternatively, you could use only filled icons and use color to represent the statistic: one color for
the proportion and the other for the remaining icons that make up the whole.
• To change the color of an icon, click on the icon and the “Format Shape” panel should pop up on the
right. Click on the paint bucket icon in the "Color" section and choose a color from the dropdown.
 Ensuring the icon is visible
➢ Make sure the icon is in front of the bar and not hiding behind the bar. This can be done in several ways:
• In the "Home" tab, go to the "Drawing" section and click "Arrange" and choose one of the options to
move the icon or graphic forward or backward.
• Click on the icon and go to the "Shape Format" tab and go to "Bring Forward" and click on "Bring
Forward". If you don't see the icon on the screen, try clicking on the chart, go to the "Shape Format"
tab and go to "Send Backward" and click "Send to Back“.

1.
1.

OR

2.
2.
 TASK
➢ Go to Sheet named "Demo 10 icons" in the Excel file "Training Dataset Day 2" and copy/paste the chart
to Powerpoint.
• Create a color legend using the female and male icons. Female is hex code #C4831E and RGB (196,
131, 30) and male is hex code #440E5F and RGB(68, 14, 95).
• Add an icon in front of the bars signifying rural as well as a source.

Proportion of women and men with no schooling in rural areas (%)

25

13
 TASK

➢ Create a pictogram in PowerPoint using the female icon to represent the statistic "4 in 5 women have a
national identification card".
➢ Add the text 4 in 5 women have a national identification card with the numbers larger than the text.
➢ Copy the finished product into the Sheet named "Demo 10 icons" in the Excel file "Training Dataset Day 2".
 What We Do
ADDING LINES OR SHAPES
FOR HIGHLIGHTING INSIGHTS
 Adding arrows for easier interpretation
➢ Arrows when combined with text can provide more context about how the reader should interpret the
scale or units of the chart.
➢ Arrows can also be used with a data callout for percentage point difference to allow readers to easily
quantify the gaps that they are seeing visually.
 Adding lines for easier interpretation
➢ Solid or dotted lines can be added as reference or baseline values. They are often added with text that
allows the reader to better interpret data values above and below the lines.
➢ Lines for gender parity are always added to signify equal male and female values in order to help the
readers understand how far from gender equality each observation is.
 Adding lines and shading to highlight insights
➢ Dotted lines, arrows, and shaded areas can help draw the readers’ eyes to gaps, outliers, ranges of years,
values, etc., and other insights more easily.

Rural men and women

have the lowest education
rates, but the urban-rural
education gap is bigger for men.

Rural men are 5 times less likely

than urban men to have any
schooling.
 Example annotation for a presentation
In Country A, 1 in 5 women have no education and the same value increases to 1 in 4 women in rural areas.
Far fewer women than men are gaining life skills from school that would improve their chances of having decent, safe, paid work.
While education rates are lower in rural areas relative to urban areas for both men and women, the urban-rural discrepancy is much
bigger for men than for women.
Percentage of the population age 15+ with no education in Country A

Rural men are 5

The gender gap is times less likely
biggest in urban areas than urban men to
where women are 2.6 have no education.
times less likely than
men to have no
education. Be consistent in the use of terminology
when presenting data. In this case, "no
education” should not be substituted
with “no schooling”, “individuals without
education” or any other terms.
 Adding lines and shading to highlight insights
➢ Dotted lines, arrows, and shaded areas can help draw the readers’ eyes to gaps, outliers, ranges of years,
values, etc., and other insights more easily.
 Arrows, dotted lines, and shading
➢ Combining the arrows with shaded areas can further highlight gaps.
➢ It is not always necessary to add a data callout in the annotation - the value of the gap can be mentioned
in the supplementary text.

National Urban Rural

 Adding shapes as annotations
➢ To add a shape, go to the “Insert” tab and click “Shapes” within the “Illustrations” section. Select the
shape you are looking for then drag the mouse on the screen to create the shape. Adjust the shape
further by clicking the dots and moving the sides.
➢ For each shape, there are options for the outline to make it thicker, dotted, change the color, or add
arrows only for lines.
 Adding shapes as annotations
➢ For shapes like a rectangle, you can choose a fill color or leave it unfilled so that there is an outline. You
can also make the fill shape semi-transparent in order to highlight a grey area or a specific light color such
as yellow or orange or red.
➢ To make it transparent, click on the shape and go to the "Shape Format" panel on the right. Adjust the
transparency with the slider in "Transparency". You can also remove the outline from the transparent
shape.
 Adding arrows as annotations
➢ Once you've made a shape, if you need to use it somewhere else, you can copy and paste it and position
it. You can copy multiple shapes at once by holding down the CTRL key and selecting multiple shapes
while holding down the CTRL key.
➢ Here we are copying the arrows, lines, and shaded rectangle for the middle visual.

National Urban Rural

 TASK
➢ Go to the Sheet named "Demo 11 Lines Shapes" in the Excel file "Training Dataset Day 2
and copy/paste the chart to Powerpoint.
➢ Add a double-ended arrow to highlight the gap with an annotation like the percentage point difference
or some other text that gives context to the gap.

Percentage without access to internet (%)

19.8

11.3

8.4

3.2

Rural Urban

Female Male
 TASK
➢ Add a shaded rectangle or outline of a rectangle to the range plot from the demo in Module 7.
➢ Copy the finished product into the Sheet named "Demo 11 Lines Shapes" in the Excel file
"Training Dataset Day 2".
 TASK
➢ Go to Sheet named "Demo 9 Text" using the chart created in Module 5.
➢ Transform the "Overall" data point in the purple column chart to a line. Add a notation that it is the
overall value.

FEMALE TO MALE EMPLOYMENT RATIO

National

Extended Household Couple

Household Couple with
Family without without
with children children
children children
 What We Do

PRACTICE EXERCISES:
 Exercise – annotate demo data visualizations
➢ Take any of the visuals we have created in the demonstrations or tasks and polish the visualization
• Add annotations (labels, arrows, shading, etc.), a title (potentially a subtitle), and source.

➢ Tips to remember:
• Annotations should use an accent color that is complementary to the visual, so it doesn’t introduce
too many colors new colors into the visual.
• Annotations and highlights should contrast well with the original visual.
• Annotations and highlights should help the reader, not distract and make the visual cluttered.

➢ Time for exercise: 20 minutes.