XKCD. If you don't know it, google it now.

XKCD. If you don’t know it, google it now.

People sometimes as me “Teddy, what exactly is your epistemology?”

To which I usually reply “LOL! That’s not a word! :D”

But jokes aside this is actually an important issue, and something that everyone should grapple with. You see friends “epistemology” basically means “how do you know which stuff is real?”

Or more precisely “what techniques or approaches do you use to work out what is real and what isn’t?”

It seems to me as if the most common epistemology is “I heard it from some guy”, which isn’t a very good one.

This can of course be extended to gossip, gossip magazines, unsubstantiated blog posts, most of Alex Jones and conspiracy theories.

Dude, thanks to Big Pharma and Big Food we don't HAVE a control group...

Dude, thanks to Big Pharma and Big Food we don’t HAVE a control group…

Although conspiracy theorists are such an interesting case that I feel they deserve a closer look.

Conspiracy theorists don’t believe ‘nonsense’, it’s a lot more complicated than that.

What conspiracy theories are about is latching onto a central idea and then scouring all of reality (or at least YouTube, which is the same thing) for any evidence that supports the idea, while simultaneously creating some kind of mental defence against any evidence that the idea is false.

The most common defence is to just ignore anything that contradicts the central idea. But this inevitably breaks down as one interacts with other people so the second line of defence is to find or make some kind of counter argument, even if it rests on the flimsiest evidence imaginable.

For example 9/11 ‘Truthers’ want you to believe in a massive conspiracy that involves Bush, the bin Laden family, the media (ALL the media!), special forces, Israel, Popular Mechanics magazine, thousands of people who lost loved ones in the attack, and every single fire fighter in New York City.

These people are capable of simultaneously believing in that laser attacks, controlled demolitions, and planes loaded with explosives, were all responsible for bringing the towers down.

But the obvious illogicality doesn’t register because it constitutes contradictory evidence and, as I said above, all contradictory evidence is ignored. And the worst part about this is that they think that they are the rational ones. “Look at all this evidence!” they will say, while waving a teaspoonful, as an ocean of contradictory evidence sloshes gently nearby.

*slosh* *slosh* *slosh*

*slosh* *slosh* *slosh*

Of course the key question then becomes “how do we do better”?

The answer may surprise you: science. Science is better at finding things out, we know this.

The simple fact that you can read this message means that science is pretty darn good at finding crap out (like how to create the most efficient lolcat delivery system ever devised).

But the thing is that ‘science’ is so ubiquitous these days that we’ve started acting as if everything that calls itself science actually is, and it isn’t. Face creams that claim to be able to influence your DNA are not science. They’re not even science-adjacent.

They are just rubbish.

Science is a set of processes intended to work out whether something is true or not and while there is a lot of debate about which methods are the best the basic format is the same:

1. find evidence to support your idea and then
2. work out if there is any reason why that evidence might be inadequate and then
3. find new, better evidence and then
4. repeat, forever. Let me use the example of homeopathy.

The question gets asked “does homeopathy work”? To answer it we first find some people who are sick, give them a homeopathic remedy to see if they then get better and what do we find? We find that they do get better! So homeopathy works, right?

High-fives all round?

Well no, sadly not. Because the problem here is that you don’t know whether it was the homeopathy that healed these people or whether it was something else. I mean hell, most of the time people get better all by themselves. How can we know whether the homeopathy was the key factor?

So we create a new design that involves a “control group”. In this experiment we divide the sick people in half, give half of them the homeopathic remedy and half of them (the control group) get nothing. Then we check whether the group that got the ‘treatment’ got better more quickly. And we find that they do! Homeopathy wins again!

High five?

Nope, wrong again. The problem now is the ‘placebo effect’. The placebo effect basically means that if someone believes that something will make them feel better then they will feel better, even if the thing they believe in has no actual effect.

There is a lot of argument about why the placebo effect works and its effects are pretty small but they are still large enough that you can’t just ignore them.

So next time we give both groups something except one of them gets the ‘real’ homeopathy and the other group just gets water.

Containing just the smallest amount of happy little ferret.

Containing just the smallest amount of happy little ferret.

But now we have a new problem because if the people doing the experiment know which group someone is in then their own biases can affect their results.

This seems counter-intuitive but it’s been proven over and over again: if experimenters know who is in the control group, and who is in the experimental group, then this influences the results. Their bias might be entirely subconscious, but it’s there and it can’t be ignored.

So now we need a ‘double-blind’ study which means that neither the participants, nor the investigators know who is getting the homeopathic treatment and who is just getting water. Then, just to be safe we get other scientists to review our work and tell us whether they can find any obvious flaws. This is called “peer review”.

And when we do peer-reviewed, placebo-controlled, double-blind studies what do we find?

We find that homeopathy has absolutely no effect whatsoever.

None. Zip. Zilch. Homeopathic remedies do not do anything.

This actually makes sense since homeopathic remedies are basically just water. And by “basically” I mean 99.999999999% water, and that remaining percent is so small that it isn’t enough to have an effect on anyone.

But now a problem arises. As you’ve seen above the process of scientific investigation takes time, and even if you start with the peer-reviewed, double-blind, placebo-controlled trials (which I will now refer to as ‘the full bla-bla’) there are numerous other problems that can creep in.

The most important of these boils down to plain old luck.

Let’s say you are testing a new anti-biotic. You do the full bla-bla and you find that not only does the new substance work but side-effects are almost non-existent.

High five?

(Concluded on Next Page)

Related Posts: