Should You Trust Science?
The scientists said we needed to eat more eggs. Then they said that eggs were bad for us–something about cholesterol. Now they say we should just eat egg whites so we can get the good cholesterol.
Who knows what they’ll be wrong about next.
The trick about science is that it’s almost always wrong. In fact, science NEVER proves anything. Any time you hear “scientifically proven” or “backed by science” you’re almost always going to get a ear full of garbage.
How can this be? Below are some more strong examples.
How to lie with statistics:
Let’s say we’re studying 4 subjects and measuring the correlation between height and how much they like using hotel showers (I don’t know how you’d get funding for this study but whatever just stick with me).
Let’s say .5 is our final correlation value. Meaning that for every unit height, we have 2 units of how much they like using hotel showers. It’s a super strong positive correlation. Who would have thought?
Now let’s say we add a 5th subject, (who would then represent 20% of the sample group). They could sway the total correlation dramatically. Each participant has a lot of “statistical weight” in the outcome because to total number of subjects is low.
If you’re following where I’m going with this, you’re realizing that every subject we add means that we get a more stable measurement. Basically, the more subjects we examine, we can be more sure that our results are good. If we have fewer subjects, our correlation value has to be realllllly high to compensate enough to be sure our results aren’t just due to chance. It’s a math thing.
This statistical concept pulls in a few greek letters that are a pain in the ass to explain (it’s called an alpha value and it’s related to beta weight,) but in a nutshell; you’ve got to monitor a ratio in order to provide statistically significant (and accurate!) results.
If you don’t want to be fooled by quoted stats, read this.
Read this if you’re poor, like I was in college. (I’ve included this in the library in my VAULT, to which you have infinite access just by being a reader).
If people are using statistics to lie, I’ll just use my brain to reason logically instead . . .
Seems legit, right? I mean, it’s at least partially true, isn’t it?
Give me a second to explain this one . . .
Your brain is incredible. It’s so amazing it’s actually a miracle. Think about it. You’ve got a 3 lb. wad of cells that can ponder the infiniteness of the universe and itself. That’s pretty impressive–but it has some shortcomings and by knowing what they are, we can protect ourselves (from ourselves).
Some of these strange shortcomings are called logical fallacies. These are erroneous reasonings and false truths that slip past us because they appear so real!
You’ve heard some of them before, and perhaps even used them! It happens all the time in arguments and it’s just not valid. It’s a normal part of being human, but it’s not useful.
Here is a slick website that goes into more nerdy detail about these kinds of common human pitfalls, and explains these logical fallacies swiftly and simply.
Science is just wrong sometimes. Actually, about 5% of the time to be exact.
It’s true. Think of it this way; when you get your study published in Emotion, a very reputable journal, you report your certainty with a p value of <.05. That means that you’re 95% sure that your correlation isn’t due to chance. This is a pretty standard requirement and most reputable journals require this standard of accuracy.
So if you ran this study 100 times, it would be statistically possible that you’d get 5 studies which inaccurately rejected the hypothesis and were still published (95 times out of 100).
Now think about all of the science that’s happening at Universities across the globe, right now. That’s a lot of science. More than 100 studies. And many of those results are getting published in reputable journals!
This DOES NOT mean that all science is wrong.
One research study is NOT science. A field of research, a collection of studies, is science.
So that cutting edge research you heard about? It’s cutting edge, but it’s not going to change any scientists’ minds until a LOT of other scientists do the same examination and get the same results.
Science only works because I write down how I collected my data, what I expected to happen, and what actually happened. I also write down all the things I think could have gone wrong with the experiment and that other scientists should look out for. Once I do this, I’ve armed others to replicate my study. Only if someone in Australia can replicate it with the same results, and then someone in Ireland can replicate it with the same results, and then someone in Uruguay can replicate it with the same results, can our findings be found valid by the field. It just means that it’s important to have a lot of studies, so that we can see the outliers and call them out.
This is important for 2 relevant reasons:
- Science doesn’t have major breakthroughs or bursts of findings. It takes a long time and a lot of smart people to figure things out.
- Science is an incredibly strict method of observation that requires years of training, patience, and persistence to achieve results.
So wait, I thought scientists were wrong all the time. How can they be smart, thorough, and consistent if they always seem to be wrong?
In fact, most people would say:
It doesn’t matter how smart scientists are, if they’re wrong then they’re wrong. If science is wrong then why should we listen to them about global warming, genetic modification, or the tenants of prenatal care?
This attitude is a great way to sadly miss the huge importance and usefulness of the scientific method.
Wrong isn’t always wrong.
Let me explain . . .
At first, scientists said the world was flat. We were wrong. Then scientists said the earth was a sphere. We were wrong again. Then scientists said the earth was an oblate spheroid. Now we’re just waiting for the buzzer, again.
But did you see what just happened? We were wrong, but how much we were wrong got smaller and smaller. To say that a child who says 4 + 3 is 576,901 is just as wrong as a child who says 4+3 is 8, is absurd! We would never say they’re the same amount of wrong!
(I’m pulling this from a great book I read in college called “The Relativity of Wrong” by Isaac Asimov, a brilliant scientist and writer)
Here’s another example:
When the first space shuttle went to the moon, it had a computer chip on board. The chip told the shuttle to point straight, move left, or shift right, depending on it’s trajectory. It would say “okay, swing a little left . . . now that’s too far, adjust right a few degrees . . . okay that’s good, now just pick the nose up a little bit . . .” until traveled all the way to the moon.
Despite it’s perfect landing on the moon’s surface, the shuttle was “off course” for 97% of the trip!
So the next time you hear someone say “oh scientists don’t know what they’re talking about,” or “scientists have just discovered x,” you’ll know better.
It’s not about one study, it’s about a field of research.
It’s not about what we see or think, it’s about what we test and galvanize.
It’s not about some big breakthrough, it’s about consistent, slow, deliberate effort.
2 Minute Action
What’s something you’re doing in your life, based on an assumption?
“I could never get into that graduate program.”
“I’m just not a creative person.”
“I’m too old (or young) to start.”
How could you test this idea? Have others felt the same as you and been wrong?
The answer is probably. So today is a great day to drop your assumptions, test your ideas and follow the evidence.
Because hey, you could be wrong.