Bad Statistics 4: Fishy Graphs

Darrell Huff’s book How to Lie with Statistics dedicates an entire chapter to the different ways in which people can lie by manipulating graphs. At first I wasn’t going to mention this because it thought it didn’t happen anymore (this book was written in 1954). But I was wrong.

A few weeks ago, astronomer Phil Plait gave a perfect example of this kind of deception in his Bad Astronomy blog. In it, he talked about climate change deniers and the “evidence” they provided to show that the planet’s climate is not increasing. These climate change “skeptics” provided the following graph:


As you guys can see, the “tenths of degree above and below 14C (world average)” stays fairly constant between 1997 and today. So there it is. Proof that global warming is completely false! … right?

Well, the thing is—as Phil Plait pointed out—these climate change “skeptics” left out a big chunk of information. If you look at the entire graph from where they obtained this data, you can see a clear difference:


As it turns out, climate change is real. What was done by climate change deniers was crop-out data (literally) that is biased to their views and presented it as legitimate. The red lining in the second graph shows the data presented by climate change deniers, while the blue lining is the data they failed to present. 

This, of course, is only one example—a real life example. This may not only happen in climate change debates, but also in other scientific areas, including psychology and behavioral sciences. This is why it is important to look for information presented by objective sources, rather than by groups and parities with an agenda, as these can manipulate data to fit their needs. 

For more information about this topic and where these graphs originated, you can follow these links: Image credit: David Rose/Daily Mail (top), Tamino (bottom).

I am not that knowledgeable on the subject of climate change, so I can’t go into great depth. The surface information will suffice as an example of how people can be deceptive with graphs.

[This is part of an ongoing series entitled Bad Statistics. Check out the previous post: Bad Statistics 3: Missing Data]

5 notes


  1. lola-forizzle reblogged this from badpsych
  2. badpsych posted this

Blog comments powered by Disqus