True False Eliminate all text to eliminate clutter

Data visualization, in the simplest terms, is a graphical representation of data to understand patterns and communicate insights.

I reckon we all are now aware of the importance of dataviz in the current times. But, we still need to focus on the essential principles for creating effective and authentic visualizations. Misleading, confusing, and impractical visualizations are common and created by even the well-known dataviz designers and journalists among us.

“There are two goals when presenting data: convey your story and establish credibility.” — Edward Tufte

Follow these five principles to create compelling and competent visualizations:

I know this sounds pretty obvious, but unfortunately, it needs to be said. There are a plethora of graphs that misguide the reader by showcasing skewed data and projecting false narratives.

The job of the presenter is to inform the audience, not influence them.

Avoid these three misleading methods to ensure that your graphs are clear and honest.

A. Omitting the baseline

Generally, the baseline for a graph should start from zero unless specified otherwise. By starting the baseline from a different number, it can bias the perception of data. This technique is used to make the difference between data points seem to be greater than actual.

B. Going against the conventions

There are certain conventions when plotting data. A larger bar indicates a greater amount, and a bigger area represents a higher number of values. By going against the convention, the viewer is unwittingly led to a wrong inference.

C. Cherry-picking data

Cherry-picking is when only a few data points are plotted to show a misleading trend. This is one of the most common tactics used to mislead or deceive the audience.

2. Know your audience

More often than not, you will be creating data visualizations to communicate information to an audience. The whole point of creating the viz is lost if the message is not conveyed.

Show what you have in a way they want to see.

A good data visualization should resonate with the audience and to make sure that it does, follow these guidelines:-

• Display data according to the job role.
• Take into consideration the education and expertise of the audience in relation to the topic of the dataviz.
• Be sensitive to ethnicity and cultural values.
• Focus on the literacy of the audience in terms of technical terms, statistics, language.

“Data is powerful. But with a good story, it is unforgettable” — Daniel Weisberg

Here, I talk about four scenarios where you should choose your data viz according to the audience you are presenting for.

A. Based on the job roles

You should present high-level data without too many details to the top-level management, while the various departments should be shown an overall picture with more details of their specific department.

B. Based on technical literacy

Use charts and figures based on the audience’s knowledge of statistics and visualization. You can show more or less complicated information and graphs according to the end-user.

C. Based on background knowledge

You must focus on the facts and numbers in the case of an audience with no prior perception of an issue. Show more details and peripheral information when the audience needs more knowledge about an issue that they are already aware of.

D. Based on emotional disposition

It is a good idea to use emotional appeal with a temperamental audience while focussing more on factual representation with a pragmatic audience.

3. Choose the right chart

Using the wrong chart is like having good intentions but poor execution.

• Choose the graph based on the kind of data and the message to be conveyed.
• Do not use different graphs just for variety, as specific graphs convey certain types of information more effectively than others.
• If not required, do not use any chart — show only numbers.

The right chart enhances information, the wrong one conceals insights.

Below, I will share some examples. For more details, check out resources and articles like this one:

Data Visualization 101: How to Choose a Chart Type

How to choose a chart type that describes your data best

towardsdatascience.com

A. Bar graph instead of a pie chart

When showing comparisons, a bar chart is generally better than a pie chart. It is easier to make out the difference in length of bars than the size of pie segments.

B. When pie charts are actually useful

Pie charts are great for showing parts of a whole, but only when there are a small number of segments and the difference between the segments is distinct.

C. Line graph instead of a bar graph

Line graphs are typically better suited than bar charts to show comparison over time. Also, line charts are very useful to indicate trends or patterns.

D. Numbers only instead of any graph

When there is a specific figure to be highlighted, it can be effective to use only numbers with no charts. One or two numbers are easy to grasp and can make a large impact.

4. Emphasize the most important facts

A data visualization can encode many data points, so highlight the most important facts to convey the message faster and with more impact.

Omit the insignificant to highlight the essential!

You can help direct attention on the most important facts by removing noise, such as unnecessary gridlines, axes and labels. Use color, size and pattern to emphasize specific data points or a focus area.

“There is no such thing as information overload. There is only bad design.” — Edward Tufte

The following three cases indicate how to accentuate what you want to communicate to the viewer.

A. Eliminate distractions

Along with using the right chart, it is important to use colors and text strategically. Moreover, if there is a conclusive result, you should enhance that in the visualization.

B. Enhance the essential

It is helpful to use the title to convey the crux of the viz and remove any irrelevant grids, labels and bright colors. You can also use reference lines and text to draw attention to specific points.

C. Highlight the significant

A distinct color or even different hue/shade can be used to emphasize important data points and highlight areas for attention.

5. Form should follow function

Aesthetics are important, but should not come at the cost of missing the point. It is more important to communicate the message clearly than to have an eye-candy graph offering no value.

Purpose of data visualization is insights, not pictures.

An intuitive design is more important than appealing charts, and graphs should convey the meaning of data in an easy-to-understand manner.

“Confusion and clutter are failures of design, not attributes of information.” — Edward Tufte

There are many such graphs — here are two that may look fun and pretty but don't successfully impart value to the viewer.

“If we have data, let’s look at data. If all we have are opinions, let’s go with mine.” — Jim Barksdale

Data is an invaluable resource and data visualization is one of the most effective tools for analysing and communicating interesting ideas and insights from the data. But badly conceived, incorrectly created or downright untruthful visualizations miss the whole point of visualizing data. Remember these basic principles to make your visualizations impactful and prolific!