User Avatar

Daniel Kubler 🚒

πŸ“ˆData management

2y ago

Ideas -> Solutions -> Results

5 Warning Signs of Unreliable Data: What to Look Out For
Daniel KΓΌbler

You definitely heard the quote about "Lies, damned lies and statistics". In the word of dashboards, this is more true than ever. Here are five clues to help you spot misleading charts and dashboards.

  1. Unclear axises. When you don't see any labels, units, or scales next to your X and Y axis, watch out. This lack can lead directly to false conclusions. Another detail to consider is whether all the charts are displayed on the same interval.

  2. Summarizing non-additive subsets. If two segments have a common part, they cannot be added. If you have 1000 Facebook and 500 X subscribes you cant report 1500 in total. There may be overlaps who reduce the total number.

  3. Piecharts everywhere! – A pie chart can look spectacular, but without the right context and detail it tells you little about your data set. You can't know the true size of a slice until you know the size of the whole data set you're looking at. A pie chart will never tell you that.

  4. Results presented only in percentages. When you don't see "nominal" values on a chart you don't see the whole picture. It's easy to report 300% growth and go bankrupt if that ratio means increasing from one guest daily to three in a restaurant.

  5. Missing baseline or benchmark data. You cannot tell if your numbers are good or bad without comparison. Data points shown alone makes this hard to decide. Yet that's exactly what a dashboard is supposed to do: tell you at a glance if there's a problem.

See what these red flags share? Each one takes something away from the context. That makes full understanding unreachable.

Pay attention to the signs and don't let yourself be fooled, no matter how nice and modern the dashboard.

The all-in-one writing platform.

Write, publish everywhere, see what works, and become a better writer - all in one place.

Trusted by 80,000+ writers