Here's one.
Let's say you compared your height to your brother's height.
You may see something more like an r of 0.7, where on
average if you have a tall brother you are likely to be tall, on average, if you have
a short brother you're likely to be short, but it's not this perfect correlation.
There's some individuals who have a tall brother but
are a little bit on the short side, and things like that.
So it's not perfectly predicted.
And again, in this case, some of this may be from genetics and
some of this may be from shared environment.
That's why you do see some sort of shared association here.
Now let's go a little bit further.
Let's imagine there's no correlation.
You may see this in the context of your height to your neighbor's height,
your next door neighbor's height.
You know it's quite possible that you'll have some very short neighbors.
You may have some very tall neighbors.
But there's no reason to expect that their height is any more or
less like your height.
So this would be a case where the correlation is 0.
And you may be wondering, well I've only done this range from 0 to 1,
what about the negative part?
Well negative is when you have literally a negative prediction such
as daily high temperature to morning rainfall in the month of April.
That you may have a lot of rainfall associated with low temperature, but
as rainfall goes down your associate you get a higher and higher temperature.
Okay, so this would be an example of a negative correlation or
a negative 0.5 in this particular example.
So I hope that gives you an illustration of what I mean when I say correlations.
So let's come back and use this.
The prediction, as I've said before,
is that if there's a genetic component to a trait monozygotic twins
should have a stronger correlation in that trait than dizygotic twins, right?
The monozygotic are the ones that are genetically identical.
Right? And dizygotic ones
have some different genes.
We're assuming that the environment is about the same in both.
So what do we see with the data?
Well if we look at correlations which ones we often see this difference
between monozygotic twins and dizygotic twins.
Here are a couple of examples.
One is for IQ, or the supposed intelligence coefficient.
The monozygotic twins tend to score very similarly on these tests.
The correlation is 0.85.
Whereas dizygotic twins have a correlation of 0.42.
This suggests there is a genetic component to what is measured here as IQ.
Same sort of thing for some diseases.
When you look at gastroesophageal reflux disease monozygotic twins have a 0.29,
dizygotic twins have a 0.13.
Again we see this higher association when they are genetically the same
than when they are genetically similar, but
not identical which means that there is some sort of genetic component.
There may be an environmental component.
I should stress this.
None of proved the absence of an environmental component.
All it shows is the presence of a genetic component.
I should stress that.