fbpx

Data Visualization: How Do We Choose the Right One?

Data Space Admin
6 Sep 23
1

In your daily work life, you will likely have to (but not limited to) prepare presentations, understand performance trends, or analyze data to find operational insights. Before creating your master piece, you will have to create visuals from various kinds of data points – date period, categories, amount, and monetary value are some common examples. But the question is, how will you know if the visualization that you have chosen is the most effective and best represents your data set?

In data analysis, data visualizations are no doubt essential for their capability to summarize large amounts of data into a graphical representation. However, to create what is considered a “good” visualization requires a certain amount of strategic planning and design thinking based on your audience and the story you want to tell. We will be exploring 4 questions that will help you select the best visualizations.

Do you want to understand the relationship between data points?

  • Scatter Plot: relationship between 2 numerical data sets
  • Bubble Plot: relationship between 3 numerical data sets
  • Heat Map: relationship between categorical and numerical data sets
Do you want to know the distribution of the data?

Histogram: distribution of 2 numerical data sets

Box Plot: distribution using statistical summary (mean, quartile, max, min)

Do you want to compare multiple items, and do you need to compare it over time? – Grouped Bar Chart: few categorical data comparisons, with same value range – Bar & Line Chart: comparing 2 categorical data, with different value range – Bullet Chart: comparing numerical from categorical data against a targe
Do you want to see how the composition looks like, and do you need to compare it over time? ·         Pie/Donut Chart: composition sliced into proportional parts ·         Waterfall Chart: composition to show accumulation or subtraction to total ·         Stacked Area Chart: changing composition over many periods ·        Stacked Bar Chart: changing composition for absolute or percentage values over short periods

The types of charts introduced here are just the commonly used ones, where there a still many more that are used for more specific purposes. Each kind of visualization has their own strong points and the use cases they can apply to. You can now clearly see how the problem statement or question affects the method that the data should be displayed as well as the message that is passed on to the audience. Keep these questions in mind the next time you have to create a visualization!

References:

https://www.datapine.com/blog/how-to-choose-the-right-data-visualization-types/
https://www.gooddata.com/blog/how-do-you-choose-right-visualization/
https://chartio.com/learn/charts/how-to-choose-data-visualization/
https://support.softwarefx.com/Chart_FX_for_WPF/article/2601233
https://etsfl.com/do-you-know-the-types-of-data/
Articles

More to Read