Richard C. answered • 07/05/16
Tutor
5
(60)
Yes, You Can Learn Math!
Since you indicate that X and Y are correlated with the same group of 10 for both values, then I'd say, yes, there's a good possibility X and Y will be correlated in the experiment.
Richard C.
Kav,
Even though you could generate a correlation coefficient for this data, it would not mean much. Typically, correlation is calculated with ONE group using two variable measures (in this case X and Y). An example would be "is there a significant correlation between age and blood pressure".
The experiment posed in the problem could be viewed as an example of what's called a correlated group test if the two random groups (but the two groups are just subsets of the original group) were measured on the the SAME variable. In the example I gave above, if a significant correlation between age and BP is found, then randomly assign the entire group to two treatment groups (A & B) and measure the effectiveness of each treatment on the dependent variable (say fatigue). We'd use a correlated t-test to do this.
So, in summary, no, I don't think it makes sense to do the correlation in this case.
Report
07/06/16
Kav S.
07/06/16