The Probability of Understanding Stastics

When my father fell and broke his hip in 2004 he’d just had his last cigarette, ending approximately 75 years of continuous and occasionally heavy smoking. His remaining years in a nursing home involved weaning off nicotine, followed by periodic searches for his cigarettes which he was convinced he had simply misplaced.

As he aged prior to dementia he would occasionally use his longevity as an indicator that smoking wasn’t really that bad for you.

I knew better than to argue with him.

But … that’s not how statistics and probability work.

I’m routinely amazed by the number of otherwise intelligent people that weigh hearsay and personal, anecdotal experience over sometimes massive amounts of more objective data.

I’ll just say I think it played a large part in recent political changes here in the U.S..

There are many issues at play.

I’m sure we’re wired to trust people and sources we don’t know more than those we do not. At the simplest level this predisposes us to believing an incorrect friend over an accurate stranger. And of course we’re almost certainly wired to trust our own experience the most.

Naturally there’s confirmation bias: we’re more likely to accept information sources confirming our beliefs. Information that disagrees has a higher barrier to to overcome.

We generally can’t comprehend very large or very small numbers. A one-in-ten chance doesn’t seem that much different from one-in-a-hundred. It might even feel like it’s not really that different than one-in-a-million. Yet the differences are dramatic, even if they don’t feel that way.

Even when we do grasp large numbers, it’s often difficult to accept that by definition every bell curve has outliers. Just because you happen to be or experience one of the outliers doesn’t mean the rest of the curve doesn’t exist. The majority of others will still fall elsewhere. My father was lucky to land where he did on that curve, but it was no absolution of smoking in general.

Objectivity is hard. Clearly the deck is stacked against us.

It’s somewhat ironic, but all we can really do is increase the probability that we’re properly factoring in our own limitations and biases.

And the first step to doing so? Recognizing the problem.

Recognize that you might not always be as objective as you think. Understand that bell curves exist and be open to the possibility that what you understand or what you experience might not actually apply to everyone.

We’ll never be perfect, but we can work towards the clear-headed and open-minded side of the cognitive bias bell curve.