Graphs and Charts and Percentages, oh my!
Source: James Vaughan
If you want to be able to find reliable information online and in the media, you will need to understand the data that these sources reference. Numbers and statistics can be represented in all kinds of ways, so it is important to know what the real story is behind those figures and to not take what is represented at face value.
How to Understand the Numbers
When reading a news article you may see a headline claiming that a program or service is spending a huge amount of money, that it’s being wasted, and that the program should be cut. It can be challenging to not agree right away with what the reporter is telling us when the number is in the millions or even billions of dollars. However, just looking at that one dollar amount isn’t giving us the whole picture.
Let’s look at a news headline, "Government program has $70 million dollars of fraud every year. Senators demand program’s end to save taxpayer dollars." Seventy-million dollars is a lot of money, but let’s zoom out and look at what portion it is of the entire program’s budget. If the program has a budget of $70.8 billion, that $70 million wasted to fraud is only 0.09% of the total amount of money spent every year. While it’s natural to be upset at government wasting money now that you see the whole picture, would you cancel the program that served tens of millions of people every year when only a tiny portion of it was reported as fraudulent?
Anytime you see an article discussing a large amount of money, see if you can find out what percentage that number is of the whole budget. Here are some websites to help you:
- USA Spending - The Federal Funding Accountability and Transparency Act of 2006 mandated the creation of a publicly available, searchable website that would provide the American public access to information on entities and organizations receiving federal funds.
- California Budget
- San Jose Budget
Causation and Correlation
Source: XKCD
Have you ever read a news headline which claimed something like, "Ice cream sales lead to drownings state wide," or "Divorce rates decline because people are eating less margarine." These articles then go onto explain that a study was conducted which shows that A (higher ice cream sales) directly causes B (an increase in drownings). While two things may or may not seem related at first glance it is not always true that one thing causes the other to occur. Again, we need to step back and look at the whole picture.
Divorce & Margarine Example
Let’s look at the divorce and margarine example. The headline "Divorce rates decline because people are eating less margarine" leads us to believe that because people are eating less margarine they aren’t getting divorced as often. Margarine must be a marriage killer! The article may even show you a graph like this one which has a line showing a decrease in people eating margarine and another line which also shows divorce rates decreasing in Maine.
Source: Tyler Vigen
What other factors do we need to consider when looking at those two sets of data? What else is going on in the environment? We know need to look at the larger story. These two items may not have anything to do with one another but the data follows the same trend which may make it look like they’re related.
Ice Cream Example
When reading a news headline which tells you that one thing caused another thing try to think if there are any other factors which are in play. If someone finds a link between an increase in ice cream sales and drownings, consider that more ice cream is sold in the summer, which is also when more people swim and unfortunately drown. The amount of ice cream being sold isn’t causing the drownings rather both events are occurring independently of one another.
Reading Charts and Graphs
Charts and graphs are often used to tell the story of data visually. A news story might be trying to show how government spending has increased over the last ten years or that deaths from shootings have dropped in a city. Not all of these visualizations are created equally and are sometimes created in a way that tells the story the audience wants to hear, not just the facts.
Changing the Y-Axis Range
Let’s look at some ways that data is represented visually. In the bar graphs below you can see the same data with different numbers in the y-axis. On the left the values on the y-axis start at 3.14 percent instead of 0 percent. This leads us to believe that interest rates have risen a significant amount since 2008. When you look at the graph at the right you can see that interest rates have remained relatively stable over the years. Anytime you see a bar graph which shows large changes over time, take a close look at the y-axis. If it doesn’t start at zero, you’re likely being mislead.
When Pie Charts Don't Add Up
Data visualizations are there in order to tell a story. Sometimes this means misrepresenting data by not following the standard conventions. For example, when you see a pie chart all the values should add up to 100 percent. In the chart below it appears as if all three candidates have about one-third of the support. That is however, not an accurate representation.
Source: Business Insider
Swapping the Y-Axis
Another example is inverting the y-axis. In the graph below it appears at first glance as if gun deaths have decreased since the ‘Stand Your Ground’ law was enacted. Taking a closer look you’ll see that the y-axis has been flipped and gun deaths actually increased after the law was put in place.
Source: Business Insider
More Resources
There are many other ways in which data is commonly misrepresented visually. Check out these resources for further reading on how you can quickly identify when you’re being mislead.
Add new comment