Skip to main content

We have a lot of data with cumbersome structures around us so analytical thinking is needed to get the things right. Some of the data makes users feel like researchers where they have to toggle brains with cups of coffee just to know the purpose of it.

How can we design the data with more comprehension?

“Storytelling with Data” is the answer.

Let’s dig deep to know how to create effective communication design systems.

Storytelling with Data helps create effective data visualization which is easier to understand. It is a visual or graphical representation of data with two purposes: 1. Communication 2. Sense Making.

Every data has some stories and it is the designer’s responsibility to get the story across to the users clearly. These are called data stories.

We can write rich and dense stories with data. We can educate the reader’s eye to become familiar with visual languages that convey the true depth of complex stories.

The process has two main sub-categories: Importance of Context and Choice of Right Visual

1. Importance of Context
If I don’t not know how to speak, I can’t communicate. Ground zero work is very indispensable to know the data set better. It includes two steps: Exploratory and Explanatory.

Exploratory
Exploratory is the research work which leads to test different hypotheses with the data and decide on a particular goal. Understand data well and figure out what might be interesting to highlight for the user.

Explanatory
Explanatory is a way of communicating the analysis derived from the research. It has 3 sub-steps:

a. WHO is the target audience or personas? It is very important to know who the user is.

b. WHAT goal or hypothesis do you want to communicate?

c. HOW do you like to communicate the hypothesis to the user in the simplest form?

2. Choose the right visual
There are many instances where I am asked to design a graph for a specific data set and I start looking for mood-boards and interactive graphs for inspiration from the word go. After a bit of search I get a beautiful graph out and then look into the data. Now I try to fix my data into the graph which is copied, oh sorry, inspired from a great design. WAIT… This is a wrong practice.

Let’s learn how to choose the best visuals for different data stories.

a. Simple Text: It is suitable when you have just one number or two that you want to communicate. Having some number does not always mean you need a graph. You can use the numbers directly. It is the fastest way to get the information across when you are dealing with one or two numbers.

b. Table: It interacts with our verbal system, easy to communicate different units of measure which deal in large data sets.

c. Heatmap: Color saturation is used to provide visual clues to help our eyes and brains more quickly target the potential points of interest. It is helpful while working with large data sets. User can quickly have a look at the critical areas which need attention. There are two types of heatmaps: 1. Numerical Data 2. Categorical Data

Categorical data is colour-coded, while numerical data requires a colour scale that blends from one colour to another in-order to represent the difference in high and low values.

d. Graph: Works with our visual system which is very fast in processing information. It means a well-designed graph gets the information across more quickly. See below the types of graph:

Scatterplot: Useful in showing the relationship between two things, because it allows you to have data simultaneously on a horizontal x-axis and vertical y-axis. It helps in showing how much one variable is affected by another. This relationship between two variables is called their correlation.

Let’s take a look at one of the data story with its graphical representation. If we are supposed to analyse the average of a group of bikes during a specific time, I would have a scatterplot graph as shown above. Each dot represents a bike while on x-axis we have miles and on y-axis cost. It shows correlation between distance driven and cost. It is very clear here that if a bike runs less than 1000 km or runs above 2000 it gives a good average. So the relation between two variables is very clear and the story is comprehensible to the user.

Line Graph: It is very useful when the story is about continuous data. Points are physically connected via line which implies a connection. It shows one series, two series and multiple series of data. It depicts trends which helps in making assumptions about the result of data, not yet recorded.

Slope graph: It is useful when you have two time periods or points of comparison and want to convey quickly the relative increase and decrease or difference between two data points. It focuses on telling the comparison between two groups. In the graph below you can see the comparison between 2014 and 2015. Team 1 and Team 2 have a positive story but Team 3 graph declined unexpectedly which is conveyed very clearly.

Bar chart: It is a diagram in which the numerical values of variables are represented by the length of lines or rectangles of equal width. It may contain single, dual or multiple series. It is the simplest graph to read. It is good to compare different statistical data of different categories. Our brain is very good at comparing things which are laid horizontally in a spatial structure.

Stacked Bar Chart: Though use cases are limited, it is a very useful graph. It allows you to compare the totals for different categories and also see the sub-components of each category. So it is a multipurpose graph. But it has some negative points as well. It is bit hard to compare sub-components across the various sub-components because you do not have a consistent base line to compare.

Waterfall Chart: It is used to pull apart the pieces of a stacked chart to focus on one at a time. It is a breakdown of a specific category which focuses on any increase and decrease in a specific category and intended result. It helps in understanding the cumulative effect of sequentially-introduced positive or negative values. It is also know as flying bricks.

Pie Chart: It is a type of graph in which a circle is divided into sectors that each represent a proportion of the whole. The human eye isn’t good at ascribing quantitative value to a two-dimensional space. It means pie charts are hard for people to read. So think twice before using a pie-chart.

Closing out

We have shown the different kinds of graphs which can be used for different data stories according to specific hypotheses. So work with the graphs, it’s fun and TELL STORY WITH DATA.

Next update coming soon with lots of more data stories and graphs. Stay tuned. Hope you liked the post!!! 🙂

Leave a Reply