byAmanda Leung

Common tricks to manipulate data:

1. Truncated Y-axis
2. Omitting Data
3. Fake Cumulative graph
4. Non-Correlating Causation

• Not setting the origin of the y-axis at zero, small differences become hyperbolic.

The interest rates are skyrocketing! And the bar sizes imply that rates in 2012 are several times higher than those in 2008.

But something seems wrong...The y-axis does not start at 0.00%.

What if the y-axis starts at 0.00% ? The interest rates are staying static now.

• By omitting certain data points, trends that don’t actually exist can easily be created whereas some existing highlights can go unnoticed.

The graph above is accurate and includes data from each year.

When half of the data points are removed, the data looks like a steady march upward.

By only plotting every second year instead of every year, the graph appears to have a steady increase, while the real data is more volatile.

• By creating cumulative graph, the data keep increasing while there is a decrease actually.

The cumulative annual revenue is moving up and to the right, so things must be going well!

When we draw a non-cumulative graph, the result is totally different...

Revenues have been declining for the past ten years!

• Correlation doesn’t imply causation, but sometimes people care more about correlation rather than causation.

Does ice-cream consumption leads to murder? NO!

We are beginning to see correlating causation more and more with big data analyses.

Data scientists are finding statistical patterns in data and sometimes care more about correlation rather than causation.

Methods to avoid the pitfalls of misleading data:

1. Using standard model for visual models
2. Make the data visualization clear and easy to understand
3. Be sure that all data and visualization has been scrutinized before it goes public