© Robert Niles
Percent changes are useful to help people understand changes in a value over time. Again, figuring this one requires nothing more than fourth-grade math.
Simply subtract the old value from the new value, then divide by the old value.
Multiply the result by 100 and slap a % sign on it. That's your percent change.
Let's say Springfield had 50 murders last year, as did Capital City. So there's no difference in crime between these cities, right? Maybe, maybe not. Let's go back and look at the number of murders in those towns in previous years, so we can determine a percent change.
Five years ago, Capital City had 42 murders while Springfield had just 29.
Subtract the old value from the new one for each city and then divide by the old values. For Capital City that means taking 50 minus 42 and dividing that result by 42. For Springfield, figure 50 minus 29 and divide that result by 29. That will show you that, over a five year period, Capital City had a 19 percent increase in murders, while Springfield's increase was more than 72 percent.
That's your lead.
Or is it? There's something else to consider when computing percent change. Take a look at a concept called per capita to find out.
Here is another way that "percent" can be misleading. Let's say that 26% of people in Springfield said last month that they liked Krusty Burger. But after Krusty Burger added its new Spicy Cluckster Sandwich this month, 29% of Springfield residents now say that they like Krusty Burger.
Doing the math, that's a nearly 12% increase in support for Krusty Burger. (Okay, here is that math: 29 minus 26 equals 3. 3 divided by 26 equals 11.54. Multiply by 100 and slap a % sign on it, and you have 11.54%, which you can round up to "nearly 12%.")
Yet sometimes people stop with the first step, which was 29 minus 26. They might say that there is just a 3 point increase in support for Krusty Burger.
That's the difference between percent change and percentage points. In this case, a 3 point difference in a poll equals a nearly 12 percent change in the public's attitude toward Krusty Burger.
Using the percent change instead of the point change provides your readers with better context about the change. But point differences can be important, too, especially when dealing with public opinion polls. We will learn more about that in Survey Sample Sizes and Margin of Error.
Read the rest of Robert Niles' Statistics Every Writer Should Know.
If these ad-free lessons helped you, would you consider sending a buck or two to Robert Niles?
© Robert Niles