The Art of Gaslighting: Statistical Sleight of Hand

The Kunlangeta Part XIII

Mar 14, 2022

"I'm a con artist in that I'm an actor. I make people believe something is real when they know perfectly well it isn't." -John Lithgow

I've got a deal for you, friend! Today is the day. There are limited stocks available.

I just got word that I'm allowed to offer you options, too. It's your lucky day. You can choose your own adventure—satisfy yourself. What more could you possibly want?

The products I'm selling today, that you get to choose from, are made by the good men and women in the most high tech pharmaceutical and biotechnology companies in the world. They're almost certain to work, and I'm told that they're very safe. But if safety is your concern, you get to choose for yourself. Don't be shy. What would you like for me to inject you with today?

Experimental Somethingorother A: No animals were harmed in the making of this particular vial, which took place after the trials in animal models. Just kidding! Ahem, where was I…oh, right…the cost is low and your insurance picks up the tab, so it's sort of like free. Can it get any better? I am obliged to tell you that the majority of the adverse reactions were serious or even fatal, but mostly just serious. Seriously, you should consider it!
Experimental Somethingorother B: Even better is that the manufacturers just rushed past the animal trials. Nothing like this product has ever been seen on Earth, so you know it's gotta be fantastic. And I'll tell you what—I chose this product and it literally saved my career. In fact, I got a promotion! Up the ladder. Imagine that. Unfortunately, I can't tell you all of the ingredients, but it contains this form of genetic code, like the kind that programs your body to do stuff. And I can't wait to find out what it does to you. I am obliged to inform you that some people who take it suffer adverse events, but more than 99.8% of those adverse events are mild and gone within 24 hours. Doesn't this sound like the better product?

Are you ready, friend? Let’s see the risks of what you’ve chosen!

Data above fabricated for the purpose of this lesson in deceptive statistics.

The Deception

Clearly, the proportion of adverse events that are mild does not determine the safety level of the product. And yet, by focusing on what doesn't matter, most everyone would have agreed from the salesman's description that Experimental Somethingorother B was the safer product.