Caution: Statistics Ahead

You’ve probably heard that half the marriages in the United States end in divorce. And so it seems, if you compare numbers of marriages and divorces in one year. Trouble is, the divorces are based on marriages that began in multiple previous years, so comparing them to the marriages in one year isn’t valid. In fact, studies suggest the divorce rate hovers closer to 30 percent and may be dropping.

That’s just one instance of misused or misinterpreted data or statistics that gained wide circulation. But for us communicators especially, it pays to make sure our usage is solidly based and not misleading.

Here are some examples of different types of slip-ups to avoid.

Spikes

A recent news report said oil and gas prices were “spiking.” This catchy word has become a staple for describing graphs of an uptick, an increase, a rise, etc. But calling such a pattern a spike hints that the number will drop, which can be misleading. Consider that over a few months starting in mid-1973, gas prices at the pump jumped from about 39 cents to 55 cents a gallon. If that had been called a spike, some drivers might still be waiting for the downturn.

Means vs. medians

In some cases, reporting mean, or average, values can disguise patterns that the median value would reveal.

For example, somebody may crow about how their efforts helped raise the average household income in an area by 50 percent. Thus, if a community of 1,000 households saw its average income jump from $50,000 to $75,000, it would appear that most made $25,000 more. But the median—the point that half the numbers are above and half below—may tell a different story. It would only take one billionaire moving into a small community to skew its average income skyward while barely budging the median.

One datum to rule them all: the myth of human error

A recent article in The Atlantic tells how in 2015, the National Highway Traffic Safety Administration declared that “the critical reason, which is the last event in the crash causal chain, was assigned to the driver in 94% of the crashes.” But it offered a key caveat: “Although the critical reason is an important part of the description of events leading up to the crash, it is not intended to be interpreted as the cause of the crash.”

But it is, and often.

The point of the article is that while the driver may own the last mistake, factors like single, easily missed speed limit signs and lanes that don’t narrow as a car approaches a populated area—which would force the driver to slow down—can set a driver up for an accident. But as long as driver error is the last in the chain of causes, that one data point tends to force all the blame onto the driver, while factors like poorly designed signage and roads and, in some cases, the heavy weights and heights of SUVs and trucks, escape scrutiny. (Heavier vehicles not only do more damage in a crash, but they are harder to stop.)

So when the 94 percent figure is quoted or driver error is cited as the cause of a crash, other data may be lurking behind the headlines—and doing nothing to bring systemic weaknesses in traffic systems to light.

If you’ve encountered other examples, please leave a comment below.