The Art of choosing the right graph. Telling a story with your data

🌟 by Jolie Pham | March 25, 2025
stacked bar chart

I still remember the first time I had to present data in a university class - my professor was there, along with a lot of my classmates. I had a beautifully detailed dataset, a thorough analysis, and what I thought was a well-prepared presentation. But when I put my charts on the screen, I could see the confusion on everyone's face. The number was there, but the story was lost. That was the moment I realized: choosing the right graph isn't just about visualising data - it's about making people understand it.

So, how do you transform raw data into a compelling visual story?

It all starts with choosing the right graph. Let's dive in!

Why choosing the right graphs matters?

Data visualisation is not just decoration - it's a crucial tool for communication. The right graph can highlight patterns, uncover insights, and make complex data digestible. What if we select the wrong graph? It can mislead, confuse, or obscure important trends.

To make the right choice, ask yourself:

  • What is the key message I want to convey?
  • What is the data am I working with?
  • Who is my audience?

By answering these questions, you can match your data to the most suitable visualization.

Common mistakes in visualisations

Pie chart overload

Using too many slices makes it unreadable. Stick to 5 or fewer categories.

Ignoring time trends

Bar charts are great for categories, but use a line chart for trends over time.

Overcrowded scatter plots

Too many points without clarity can obscure insights—use transparency or clustering.

Unlabeled axes

Always label axes and provide context to prevent misinterpretation.

Inconsistent scaling

Misleading scales can exaggerate or minimize trends. Be mindful of axis manipulation!

Types of data and their best graphs
Comparisons
Use these graphs when you need to compare values across different categories.
  • Bar chart
  • Stacked bar chart
Trends over time
When showing trends or patterns over time, focus on time-series graphs.
  • Line chart
  • Area chart
Proportions
If you want to highlight how a whole is divided into parts, consider these visualizations.
  • Pie chart
  • Doughnut chart
  • Stacked bar chart
Relationships between variables
If you want to explore relationships between two or more variables, use scatter-based charts.
  • Scatter plot
Distributions
If you want to analyze how data is distributed, use one of these charts.
  • Histogram
  • Box plot
  • Violin plot
Make your data speak!

Choosing the right graph is crucial because it can transform data from numbers to valuable insights. When selecting a visualisation, always prioritize clarity, accuracy, and storytelling.

Next time when you are about to create a chart, take a moment to ask:

Is this the best way to communicate my message?