Question

I do not get this. (>.<)

Answer 1

The correlation coefficient for poor health and not exercising in a group of people is 0.67. Analyze the following statement.

Poor health is caused by not exercising.

Is this a reasonable conclusion?

Yes; everyone who doesn't exercise is by definition unhealthy

Yes; the correlation coefficient is above 0.5, so that implies causation

No; the data are only weakly correlated, and many people are healthy who don't exercise

No; poor health and not exercising are completely unrelated

Answer 2

either b or c...

Answer 3

What's not to get? Correlation is a tricky animal. It means ONLY that things move together in a somewhat predictable fashion. If you see the word "cause", run screaming. Stay away from that. You need an experimental design for that.

Funny, some fields of study laugh at correlation as low as 0.95 That's just not good enough. Others, perhaps social sciences, have a party if correlation gets as high as 0.40. Is 0.67 "weakly correlated"? Maybe. It depends on who is talking.

This correlation of 0.67 leads to a COMMON conclusion that there is causation, but it is not a REASONABLE conclusion. There is no such data.

Answer 4

Given the four choices, I would evaluate them like this:

Idiotic
Absolutely Not
"Weakly" is subjective, but the last part is a very good point.
No one believes that! There is plenty of evidence to the contrary.

I'd go with the clear choice, C.

Answer 5

Thank you both - sorry for the late reply, though. I was devouring supper.