Thread: How do scientists determine the independent and dependent variable?

For example take the hypothesis "people who listen to Mozart tend to be intelligent". Is the dependent variable intelligence or if a person listens to Mozart? For the sake of the study is it arbitrary? I say this because I've come across 2 definitions of independent variables:
1)The variable that influences the dependent variable
2)The variable that testers manipulate
So in a situation like this it's arbitrary if a tester looks for people who do or do not listen to Mozart, or alternatively, look for people with a particular level of intelligence.

Originally Posted by superdude

For example take the hypothesis "people who listen to Mozart tend to be intelligent". Is the dependent variable intelligence or if a person listens to Mozart? For the sake of the study is it arbitrary? I say this because I've come across 2 definitions of independent variables:
1)The variable that influences the dependent variable
2)The variable that testers manipulate
So in a situation like this it's arbitrary if a tester looks for people who do or do not listen to Mozart, or alternatively, look for people with a particular level of intelligence.

I'm not familiar with the word "tend" when it comes to a logical relationship. i could be a partial "if then" or a "if and only if" relationship
I would probably rewrite the statement, but it might lose meaning.

(? not very good at Logic but I'll try)

M - listen to Mozart
I - are intelligent
x - people

For most x, if M then I

Originally Posted by MacstersUndead

I'm not familiar with the word "tend" when it comes to a logical relationship. i could be a partial "if then" or a "if and only if" relationship
I would probably rewrite the statement, but it might lose meaning.

(? not very good at Logic but I'll try)

M - listen to Mozart
I - are intelligent
x - people

For most x, if M then I

I believe you missunderstand my question. When I used the word "tend" I used it in a coloquial mannor. My point is, if research is being done on something that has never been studied before, how can a mathmatician know which is the dependent and which is the independent variable? For example "are dogs owners healthier because they own dogs" or "do healthy people own dogs because they are healthier"?

Originally Posted by superdude

I believe you missunderstand my question. When I used the word "tend" I used it in a coloquial mannor. My point is, if research is being done on something that has never been studied before, how can a mathmatician know which is the dependent and which is the independent variable? For example "are dogs owners healthier because they own dogs" or "do healthy people own dogs because they are healthier"?

It is the way you design the experiment, the variable/s you control is/are by definition the independednt variable/s, the ones you observe are dependent or extraneous.

Control in here may mean you select a subsample with that charteristic rather than dial the value in on a knob.

By the way your example is poorly choosen as all you will do is demonstrate a correlation between dog ownership and health, and "correlation is not causation"

CB