Bad Data (and Good Data Turned Bad)

(Provided by "worker" from "openclipart.org" created 2015-12-08)
Scientific data is represented and called upon a lot in modern times. Everywhere from books to internet articles to TV news stations will use graphs, statistics, and any other legitimate enough looking data to provide their audience with information they want them to know.

However, there are problems with this supposed "scientific data" that people can often miss...if it's presented to them in a certain way.

Let's start with the example of Climate Change, and a good graph from a credible source to back up the claims that human caused climate change is steadily and dangerously warming the planet.


Figure 1
(Data source: Reconstruction from ice cores. Credit: NOAA. Found on climate.nasa.gov)

Figure 2

(Data source: NASA's Goddard Institute for Space Studies (GISS). Credit: NASA/GISS. Found on climate.nasa.gov)

These two graphs provide clear, clean, and simple data and show an obvious trend in not only the rising levels of CO2 in the atmosphere (Figure 1), but the rising temperature as well (Figure 2), Figure 2 even being taken from a yearly updated and currently live feed from Nasa's Goddard Institute for Space Studies.

These two graphs are reliable. They are provided by a scientific institution that uses pure numerical data to represent their information.

However, not all graphs are given in such scientific and clear ways.

Take for example, the graph below.

Figure 3

( Posted to Twitter by National Review (@NRO) at 4:36 PM - 14 Dec 2015)

Just look at how little the average global temperature has actually changed! (according to the graph) Guess we won't ever need to worry about anymore.

While obviously Figure 3 a satire of graphs used to show climate change, the graph itself is still a perfect example of what is considered "Good Data Turned Bad." Inputting the average annual global temperature data onto a graph is still technically showing the data, but manipulations can be made to that graph to make it appear as though the data isn't significant. In this example, National Review scaled up the Y-axis of the graph (temperature in degrees Fahrenheit) to a massive level, beginning at -10 at the bottom and working up to 110 at the top. Of course, on this scale, the temperature change is going to be an essentially flat line. However, if one were to scale down the Y-axis, there would be a very clear, and even drastic, upwards trend between 55 and 60 degrees from 1880 to 2015.

All in all, it's not a bad thing to trust data. Numbers can't lie to you. But, you should be on the lookout for any data that may be represented wrongly, or by a potentially biased source that wants you to believe what they're trying to tell you. Learning to spot the difference between good and bad data can help you gain more and better insight and knowledge for the future.