Percent change in a value tells you only part of the story when you
are comparing values for several communities or groups. Another
important statistic is each group's *per capita* value. This
figure helps you compare values among groups of different size.

Let's look at Springfield and Capital City again. This year, 800,000 people live in Springfield while 600,000 live in Capital City. Five years ago, however, just 450,000 people lived in Springfield while 550,000 lived in Capital City.

Why is this important? The fact that Springfield grew so much more than Capital City over the past five years could help explain why the number of murders in Springfield increased by so much over the same period. After all, if there are more people in a city, one might expect there to be more murders.

To find out if one city really is more dangerous than another, you
need to determine a *per capita* murder rate.
That is, the number of murders *for each person in town*.
(That's what "*per capita*" means. It's Latin for "for each head.")

To find that rate, simply *divide*
the number of murders by the total population of the city. To keep
from using a tiny little decimal, statisticians usually multiply the
result by 100,000 and give the result as the number of murders per
100,000 people.

In Springfield's case, 50 murders divided by 800,000 people equals a murder rate of 6.25 per 100,000 people. Capital City's 50 murders divided by 600,000 people equals a murder rate of 8.33 per 100,000 people.

Five years ago, Springfield's 29 murders divided by 450,000 people equaled a murder rate of 6.44 per 100,000 people. And Capital City's 42 murders divided by 550,000 equaled a murder rate of 7.64 per 100,000 people.

In the previous section, we found that the number of murders in
Springfield increased 72 percent over five years, while the number of
murders in Capital City grew by just 19 percent. But when we now
compare *per capita*
murders, Springfield's murder rate decreased by almost 3 percent,
while Capital City's per capita murder rate increased by more than 9
percent.

There's the real story.

Remember in Chapter 6 [of *How to Make Money Publishing Community News Online*], when I wrote about school test scores?
That's another example of how reporters can miss stories when they
don't make apples-to-apples comparisons with data. It's just not fair
to call schools serving poor students "failures" compared
to schools with mostly of wealthy families, when family poverty is the
number-one factor affecting test scores. You've got to account for
family income when comparing student test scores across schools and
districts. In the same spirit, figuring crime stats, economic data
and other community characteristics as *per capita* numbers,
instead of using the raw numbers of incidents, can help you make the
apples-to-apples comparisons that allow you to report truthful information to your readers.

Read the rest of Robert's statistics lessons for people who don't know math.

© Robert Niles