73% of Statistics are Fake

We’ve all heard or read them before:

10,000 hours to get good at something.
23 minutes to refocus after an interruption.
10,000 steps a day are required for optimal health.
Your BMI must be between 18 and 24 for you to be healthy.
Half of the studies you read about in the news are wrong.
1% improvement everyday will lead to greatness.

The list could be extended to easily include exactly 12,386 items.

Particularly in these COVID times, the news begins to look and sound a lot like a stock market ticker tape as stats are flung around with reckless abandon.

Someone — maybe the Prime Minister of Great Britain, Benjamin Disraeli — said, “there are three kinds of lies: lies, d*** lies, and statistics.”

We can generalize the problems with stats in the following ways:

Disproportion and Amplification. When a popular enough group takes hold of a particular statistic — validated and verified or otherwise — other groups tend to get on the bandwagon. This is the danger of parroting: blindly repeating something without any fact checking. Before long, the misinformation is propagated. In the case of invalid information, the rapid propagation makes it difficult (often impossible) to get the correct information out. Having first position to the eyes and ears of the masses makes it difficult to unseat bad information even when the credible information finally comes to light.
Skewed Results. There are so many stats about studies. “In a new study published by. . .” is generally how the story starts. The news will never tell you about the study group. Those groups are often skewed themselves. For example, for decades studies on blood pressure medications excluded most ethic groups. And, as it turns out, some blood pressure medications are ineffective in large segments of the population as a result. However, their clinical studies showed enough efficacy to get them approved by the FDA. Fen-phen anyone?
Selective Use. This is the slice-and-dice method. Loads of stats have been published for and against so many things. The buffet of available statistics can be picked, a la carte style, to fit the needs of the story you’re trying to tell. Even when the stat is reliable, it doesn’t tell the entire story.

So, should we show statistics the door and make due without?

No, but there are a couple of things that you should keep in mind:

Is the statistic something that is going to impact your own health and well-being? If not, you probably don’t have to invest time to validate it. But, don’t spread it around if it will impact the well-being of others. That is, if it looks like a stat meant to incite fear, panic, or some kind of call to action, it’s best to keep it to yourself.
If it is something you need to rely on, it’s essential to know from where it came: is the data source reliable (hard to determine), is the analysis credible (again, hard to determine), or is the stat just being parroted by everyone as click bait?

The truth is out there. You just have to look hard to find it.