[Question] Valid inferences for Correlation?

alexander

Hi guys... Maybe these are obvious and I'm missing something but I couldn't really find an answer online so I was hoping someone can help me out with some general questions regarding correlation.

If given a statement of positive correlation such as "as A increases, B increases," how much can we validly infer?
Can we infer all of...
-as A dec B dec
-as B dec A dec
-as B inc A inc

Also for negative correlation -"as A increases, B decreases"
Can we infer all of...
-as A dec B inc
-as B inc A dec
-as B dec A inc

Thanks in advance

alexander

Is correlation basically like probability? and that's why you can't diagram it as conditional statements?

A = time studying
B = grade received

original statement: as A increases, B increase:
-as the time you put into studying increases, the higher your grade
is it implied...that:
-as the time you put into studying increases, the chance that you will receive a higher grade increases and NOT that your grade WILL increase? and therefore not diagrammable as a conditional statement?

and if that's correct, these statements below can be inferred correctly as...
-as A dec [it is likely that] B dec
-as B dec [it is likely that] A dec
-as B inc [it is likely that] A inc