Without evidence of benefit, an intervention should not be presumed to be beneficial or safe.

- Rogue Medic

The Difference Between a Scientist and a Normal Person


Randall Munroe (xkcd) demonstrates why he may be the poet laureate of science.

We may avoid danger by superstitiously avoiding circumstances similar to those that surrounded something bad . . .

xkcd – The Difference. Click on the image to make it larger.

How could you choose avoiding a little pain over understanding a magic lightning machine?

. . . but we also may do the opposite, when we assume that the improvement of a person after something we did was because of what we did and that unintended consequences will not result from our actions.

Rather than the person pulling the lever, what if this were a doctor, nurse, medic, EMT, . . . giving something to a patient?

The treatment is given.

The patient appears to improve.

Even though the treatment is one of many things that might have led to the patient’s apparent improvement, we assume that what we did caused the apparent improvement.

The normal person may decide that this is a beneficial treatment and give it to everyone the normal person thinks is similar to the original patient.

If a small percentage of patients appear to improve after the treatment, we remember them as successes because of the treatment.

Does this have the potential for unintended consequences?

Maybe we justify our rush to treat patients with something we do not understand because of some rationalization with a sciency sounding physiological explanation. this is probably not what is really going on, but based on remembering the examples that confirm our biases and forgetting those that do not confirm our biases, we can twist what is in the physiology text books to appear to support our treatment.

After all, we mean well, so nothing bad should happen.


That is the way the world works. Unintended consequences of best laid plans poorly understood actions never, ever go awry.

You are depriving patients of a breakthrough treatment!

It should not be difficult to provide good evidence that a treatment that good really is that good.

You are holding back a modern Galileo!

Galileo repeatedly provided excellent evidence that he was right and that Aristotle was wrong.

Galileo was a scientist. If Galileo had been a normal person, we would not learn about what he did.

The scientist recognizes how little information this anecdote provides.

The patient received one treatment, but may have been receiving many other treatments, or may have already been getting better, or may have been misdiagnosed, or . . . and that one treatment may have had nothing to do with the patient appearing to get better. That one treatment may have actually interfered with the patient getting better, but we tend to see the result we want. The patient appeared to get better.

The scientist recognizes the ethical obligation to the patient to not unnecessarily harm the patient, so the scientist tries to find out what is real.

The scientist does not look for rationalization for biases, but sets up conditions to minimize the influence of biases.

The scientist requires good evidence that the treatment works.

In this case, the normal person is an anecdotalist and is only concerned with the possibility of missing out on a good thing, but is completely ignorant of the unintended consequences of plans based on anecdote.

Avoiding the harm of the shock, based on anecdote, is not unreasonable.

Assuming the benefit of a treatment, based on anecdote, is unreasonable.