Given that A and B are either the independent variable or the dependent variable:

1) "A is correlated with B"
For this wording, does it mean that A is the independent variable and B is the dependent variable? (i.e. A affects B?) Or is it the other way around?

2) "A is correlated to B"
Is this meaning exactly the same thing as "A is correlated with B"? Or does it mean that A is the dependent variable and B is the independent variable? (i.e. B affects A?)

3) "There is a strong postive linear correlation between A and B"
"There is a strong postive linear correlation between B and A"
Is there any difference between the two statements above? Does the order of stating A and B matter here? Does this wording give any clue to whether the first one or the second one is the independent variable?

4) When it says "something vs something", should the independent variable listed first or second?


Can anyone help me, please? I am really confused now!
Thank you! :)

Numbers 1-3 have vague distinction in general practice.

Under most occasions, usually the horizontal axis is labelled as the independent variable, and comes up as the first 'something' in something vs. something. But whether it is convention, I do not know.

There are a few exceptions, such as the stress-strain graph in materials science, where stress is the independent variable that happens to appear as the ordinate.

Under most occasions, usually the horizontal axis is labelled as the independent variable, and comes up as the first 'something' in something vs. something. But whether it is convention, I do not know.
They seem to do it the other way around in economics (Price vs. Quantity). Non-conformist REBELS...

"A is correlated with B"
"A is correlated to B"
"There is a strong postive linear correlation between A and B"
"There is a strong postive linear correlation between B and A"


All these statements mean the same thing.
None tell anything about the causal relationship (if any) between the variables.

This sounds like some of the nit-picky stuff they do in lower math. But I will give it a go anyway.

Note that I am not sure without seeing what your textbook has to say about it.

#1...I believe they mean A can be written as a function of B, that is B is the independent variable. Note that it does not say "linear correlation", which is different from #3.

#2...This would be the reverse of #1. Here, A is the independent variable.

#3...Yes, these are the same thing. You can write the relation as aA+bB=c, with a,b,c being constants, and since it is a positive linear correlation a/b is negative.

You can then rewrite this as...

B=(c-aA)/b with A the independent variable, or

A=(c-bB)/a with B the independent variable.

I hope this helps. If you need an explanation of implicit functions, just ask.

Numbers 1-3 have vague distinction in general practice.

Under most occasions, usually the horizontal axis is labelled as the independent variable, and comes up as the first 'something' in something vs. something. But whether it is convention, I do not know.

There are a few exceptions, such as the stress-strain graph in materials science, where stress is the independent variable that happens to appear as the ordinate.
But for example in physics, position-time graphs, velocity-time graphs, it is y vs x (y against x) and the dependent variable is listed first....I don't know if this is the correct form......

All these statements mean the same thing.
None tell anything about the causal relationship (if any) between the variables.
I'd agree with this, since the definition of correlation I'm familiar with is symmetric.

But for example in physics, position-time graphs, velocity-time graphs, it is y vs x (y against x) and the dependent variable is listed first....I don't know if this is the correct form......
It usually is. Its just an arbitrary convention though, not one of the Ten Commandments. Don't go losing sleep over something like this...

But for example in physics, position-time graphs, velocity-time graphs, it is y vs x (y against x) and the dependent variable is listed first....I don't know if this is the correct form......

It is certainly the most common form - although you can do the reverse as well. Like time as a function of position, etc. What's correct depends on what variable you test, the dependent variable. For a set/given/controlled value of x, you plot y.