FROM DATA TO STORIES
Data, visualisation and narrative
In this lesson, you learn how to construct a data story. You will also learn effective
approaches for visual data storytelling and how to engage your audience to inspire
them to take action.
In this lesson, you will learn:
• How to construct a data story
• The fundamentals of visual storytelling
• How to inspire audiences to action
• Build a narrative that relates to an audience
Consider this example
In the visualisation above, created for the Swamp City project, we see the value and
the benefits of wetlands in Kampala, Uganda that has been modified, mainly due to
industrial development and small-scale farming (Read the story abou the Swamp
City Project using this link: https://stories.newvision.co.ug/swamp-city/ )
Storytelling is a natural element of human communication and interaction. Data
stories share much in common with other kinds of stories that we encounter every
day.
The difference is that data storytelling centres around data and its presentation.
Data stories have three important elements: data, visualisation and a narrative, as
explained below
What is data storytelling?
Data storytelling is the practice of translating complex data analyses into
understandable and engaging narratives. It combines data, visuals, and a narrative
to help the audience understand the insights hidden in the data.
It is important because it makes data more accessible, helps communicate insights
clearly, and allows decision-makers to take informed actions
Data, graphics and narratives are the main components of data storytelling, an
organized method for conveying insights gleaned from data.
When the story is supported by data, it helps the audience understand what is
going on in the data and why certain insights has been produced
when applied to data, graphics can reveal insights to the audience that they might
not have noticed without the charts or graphs
A data story can impact and motivate change when appropriate graphics are
combined with the appropriate facts
Data storytelling provides deeper insights and supporting evidence through
graphs and charts.
Through data storytelling, complicated information is simplified so that your
audience can engage with your content and make critical decisions quicker and
more confidently.
Constructing a data story that moves a person to take action can be a very
powerful tool. Effective data storytelling can have a positive impact on people and
your organization.
Data for storytelling- Components/elements of data storytelling
There are three elements of a data-driven story: the data, the visualisation, and the
narrative. This compelling image created by PesaCheck manages to combine all
three in one graphic, explaining if it's possible for cyclones to hit Kenya (read
PesaCheck story using this link here: https://pesacheck.org/is-kenya-likely-to-be-
hit-by-a-cyclone-912c07ce85ae)
• Data
Data is the foundation of any data story. It represents the raw information collected
through various sources such as surveys, experiments, or digital tools.
The data provides the factual backbone for the story and supports the insights the
storyteller wants to convey.
The first step in data storytelling is analyzing the data to uncover meaningful
patterns, trends, or anomalies that are important to the intended audience.
• Visuals
Visuals are critical for making data accessible and understandable. Visualizations
like bar charts, line graphs, scatter plots, or infographics help represent large data
sets in a simplified, easy-to-read manner.
They enhance clarity and allow the audience to quickly grasp key insights. Visuals
also help highlight patterns, relationships, and outliers that may be difficult to spot
in raw data. Well-designed visuals can make data more engaging and memorable.
• Narrative
The narrative is the element that ties the data and visuals together. It provides
context, explains the significance of the data, and leads the audience through the
story in a logical and structured way.
A good narrative answers the "why" behind the data and guides the audience from
understanding the data to realizing its implications.
The narrative is crucial for making the data relatable, helping the audience
understand how the information impacts their decisions or the subject being
discussed
Remember that data analysis starts by developing a question that you want your
data to answer. This question will provide the centrepiece of your data story.
You can think of your research question as being the scene of your data story, and
the data itself as the characters that bring the story to life.
Why tell stories with data?-Benefits of data storytelling
Data by itself is often not very compelling or interesting and can be hard to
remember. But stories command our attention.
A good data story explains how data is relevant to an audience, and what they
should do about it.
Data can make storytelling more effective because:
• It’s persuasive
Data presented in story form is more compelling than numbers alone. Data can
provide support for other evidence (such as an expert comment or anecdotal
evidence).
• Data can raise awareness
Data stories can highlight important and compelling issues that might otherwise
get lost in a sea of information.
• Data stories can foster engagement
Data-driven storytelling can make data more interesting, relevant and personal.
One can also ask readers to share their own experiences as data, such as in this
example from OpenUp, asking people to share information about what they pay
domestic workers.
• Data is a call to action
Data stories can compel an audience to take action, or adopt a different point of
view.
As you conduct your data analysis, look for characteristics that might help develop
your narrative, and provide interest.
In the interactive visualisation above, you can see some of the characteristics in a
dataset that might help develop your story.
This Valentine's Day special by OpenUp in South Africa documents the process of
finding a story in data perfectly (Read the story using this link:
Valentine's Day Edition - Love knows no age Example
You often learn a lot about people by looking at data.
According to StatsSA in the dataset of marriages and divorces, there were roughly
161,000 civil marriages in 2012.
A quick histogram of the wedding month shows that December is the most
popular month for getting married. No-one wants to get married in winter (May-
Aug)
Things start getting weird when you look at the age differences between marital
partners. Have a look at the graph below.
Every dot represents one or more marriages. As you move up the graph, the age of
the bridegroom increases. As you move to the right, the age of the bride increases.
It seems that the majority of all marriages are between people who are more or less
the same age. There are however a significant number of marriages that don’t fit
that mould.
Let’s define a sugar mommy or sugar daddy as someone who is married with a 20-
year or more age difference between them and their Ben Ten/Sugar Baby.
All the dots outside those diagonal lines in the graph represent sugar marriages.
Note that there are far more sugar daddies than sugar mommies. In all, 3,400 sugar
marriages.
There are some odd unions. Have a look at the dot on the bottom right.
It represents a marriage between an 86 year-old woman and a 16 year-old boy.
There are also examples of very old men marrying younger women. The most
disturbing part of the diagram is on the far left.
Girls younger than 16 are getting married. Two 12 year-olds were married off, one to
a 20 year old man, another to a 67 year-old. Is this even legal? Turns out that it is.
The dot on the top right between the 92 year-old bride and 94 year-old groom is
heartwarming. It’s clear that you can still find your soulmate, even after 9 decades
of searching.
Patterns and trends
How does your data change over time? Are conditions getting better, worse, or
shifting from one category to another? Are these changes connected to policy
decisions or other events?
Similarities, contrast and outliers
In what ways are groups within your data similar or different? Are some groups
performing above or below average? Are there outliers that are substantially larger
or smaller?
Interesting or unusual exceptions
What are the individual values that fall outside the norm? What might explain
these exceptions?
Data visualization-Data story visualisations
Data visualization plays a crucial role in storytelling by transforming raw data into
visual formats like charts, graphs, and infographics.
These visuals help in simplifying complex data, highlighting trends, and making the
information more accessible to the audience.
Good visualization supports the narrative and makes it easier to convey the story.
Visualisations are a key part of telling
a data story.
They let your audience see what the
data conveys and lets them explore
the data on their own.
The visualisation chosen should
match the question you are using to
create your story and the
characteristics you discovered in
your analysis as seen in the above
image published in by PesaCheck
(image given alongside)
Narrative trends over time
Data stories often describe how conditions change over time.
These stories can show the impact of long-term trends and highlight external
changes that might impact those trends.
Cases of Waterborne
Diseases in Busia
County
Between January 2019
and April 2020, 93,463
cases of waterborne
diseases were reported
in Busia County in
Kenya.
Diarrhoea accounted for
78% of all cases reported
and Typhoid Fever
accounted for 18% of all
cases.
For example, this visualisation shows the change in the number of waterborne
disease cases reported in Busia County in Kenya over the period of January 2019 to
April 2020.
Surprising or counterintuitive data
Data stories can highlight surprising or counterintuitive results.
A data story may provide evidence that contradicts and challenges common
misconceptions.
For example, the highest prevalence of female genital mutilation cases recorded in
Nigeria could be believed to be among those with a low level of education.
However, the data shows that female circumcision was more prevalent among
women with secondary or higher education levels.
Comparisons
Data stories can make comparisons between different groups or categories, and
provide context to put information into perspective.
For example, the visualisation above gives a comparison of sexual assault victims
by age, recording the changes in cases over time.
Mirabel Sexual Assault Referral Centre in Lagos, Nigeria reports that most victims
are aged 14-17 years followed by ages 6-10 years and 11-13 years.
From details to the big picture
Data stories can provide a larger context for a smaller area focus or vice versa.
This approach is especially helpful for comparing local data to a larger geographic
area.
Maps that allow users to explore their own states or neighborhoods can be
particularly effective.
In this visualisation alongside
for a story addressing
Inadequate PPE supply, the
number of doctors, nurses,
pharmacists, and community
healthcare workers in Africa
are mapped.
Readers can hover over an
area to see details in any
country.
Intersecting or diverging data
Data stories often unfold when two or more sets of data intersect or diverge.
Intersecting data refers to an instance when two or more sets of data meet and
cross at a point in time.
On the other hand, diverging data in an instance where two or more sets of data
take different directions.
Your audience will want to know why these changes occur. Common examples
include political opinion, life expectancy, and wages.
This visualisation shows
the relationship between
the number of people
affected by key natural
hazards in Kenya over
time, using data from
the World Bank Climate
Change Knowledge
Portal.
Natural hazards such as
epidemic and drought
intersect between 1991
and 1994.
Drought diverges from
the rest of the natural
hazards from 1997 to
1999.
Read the World Bank Climate Change Knowledge
portal story here https://climateknowledgeportal.worldbank.org/
Tracing Flows
Data stories can
convey complex
data relationships
that would be
almost impossible
to make sense of
otherwise.
This is especially
true for flows of
commodities, such
as goods, money,
or information
between multiple
sources and
destinations.
This example showcases aid inflow into Zambia from 2000 to 2013, using records
from AidData. It makes it easy for audiences to see the aid between particular
countries and regions, and if one hovers over a particular flow line they can get
information on the country and amount of aid donated.
