Misleading Statistics, Part 2 -

Images used under Creative Commons license from NASA The above graphs of the exact same set of data tell very different climate stories. The top left graph shows a general decline in global temperature and the other two show a significant increase in global temperature.

Images used under Creative Commons license.The above graphs are based on the same statistical data showing the change of direction of Planned Parenthood. The top graph is meant to show the shift from preventative medicine to abortions, and the bottom is meant to show that there has not been a shift in focus. However, one can see that there has been a steady decline in preventative procedures by Planned Parenthood and a steady path in abortion procedures. Neither graph in fact tells the truth about the organization because both are biased in one way or another.

By Scott Hamilton

Last week I talked about how the data could be manipulated to create statistics that can mislead people to make an incorrect decision. It is important to note that this is not always intentional, but it does happen. This week I want to continue that discussion and cover how even accurate data that has not been manipulated in any way can be used to mislead people, by the way the data is presented.

The first of these methods is used quite frequently; it is called misleading data visualization. When you are presenting statistics, you get to choose the type of charts and graphs used to represent the data. One of the top ways to manipulate charts to tell the story you intend is to manipulate the scale of the data. A great example is to use two different scales when comparing data so that you get the data lines to cross and appear to have more impact. A great example of this method was used in attempts to defund Planned Parenthood, by showing that the funding was being used primarily for performing abortions rather than for preventative medicine. You can see these graphs in Figure 1.

The second method of manipulating the charts to tell your story is to shorten the timespan of the data. This is used frequently when talking about climate change. Those that support climate change will use a long-range graph of the global average air temperature starting in 1980 and running through 2012. This chart will show the global temperature with a significant increase on average of about 0.5 degrees Celsius. Even longer timeframes show greater temperature change from 1900 to 2012 of 0.7 degrees Celsius. Those trying to disprove climate change will use a graph starting in 1998, the hottest year on record due to a strong El Nino effect, which will show a global temperature decrease of about 0.1 degrees. You can see both of these graphs in Figure 2.

The last way that statistics can be used to manipulate the public is in the form of advertising. In 2007, Colgate was advertising that “80% of dentists recommend Colgate.” This was a true fact, but deemed as illegal advertising because the public read it to mean that 80% of dentists recommended Colgate over other brands, but that statistic was actually stating that 80% of dentists would recommend Colgate as a quality toothpaste, and may also recommend any other brand of quality toothpaste at the same rate, making the statistic mean that Colgate is safe to use, but not necessarily better than any other brand.

Just like I stated last week, it is worth the effort to research the background of any statistic used to persuade you to make a choice. You need to consider the source of the data, the agenda behind the presentation, and the information presented, as it may be presented in a slanted way in order to misguide your decision. Until next week, stay safe and learn something new.