We continue the topic of data visualization. In this second article on misleading (intentional or not) visualizations, we take a closer look at two key areas — units and perspective. Why? Because data visualization is one of the key components of effective information delivery — not only in business. And at the end… a real gem.

We’ve divided the most common mistakes (or manipulation tactics) into several categories. Let’s look at them one by one.

1. Units

With the same dataset on the table, you can visualize it in entirely different ways — leading to totally different conclusions. Below you’ll see the same data: on the left shown as a dynamic index (1990 = 100), on the right presented as a share of GDP. Same world, but two very different stories.

Even trusted sources like The Economist can get it wrong. On the right, the correlation doesn’t seem so perfect. Always be cautious of charts with dual axes!

More on that topic here.

2. Perspective

Who is really the market leader in smartphones? The right chart perspective, font size, order, and color choices make it easy to believe it’s Apple… even if it’s not.

3. Hidden Trend

Here’s an example (from a great post linked below) of a chart that conveniently skips over a major trend — the decline in Beta. By carefully choosing the chart type and data order, you can radically change the message. The second chart reveals the “naked” truth.

These are just a few examples pulled from the ocean of visualizations that surround us daily. To wrap up, we’ve included some additional charts with comments on what could have been improved.

Example of poorly applied scale:

Reversed axis — presented in a counterintuitive way:

Poor chart selection — using a pie chart instead of a more readable bar chart:

Using projected data to imply a trend — without clearly marking the forecast period:

Dual axis with mismatched scales — potentially misleading in presenting COVID case data in Poland vs other countries:

If you enjoyed the last chart, you’ll find more “creativity” on: Viz.wtf

Since data visualization plays such a vital role in our work, future blog posts will showcase good practices and well-designed charts and dashboards.

Sources and useful links:

• Venngage.com
• StatisticsHowTo.com
• PraktykaTrenera.pl
• Viz.wtf