Refer to Exercise 12.31. Suppose that a completely randomized design is employed for the
Chapter 12, Problem 12.33(choose chapter or problem)
Refer to Exercise 12.31. Suppose that a completely randomized design is employed for the comparison of the two treatment means. Then, a response could be modeled by the expression Yi j = i + Pi j + i j, but the pair effect Pi j (which will still affect an experimental unit) will be randomly selected and will likely differ from one of the 2n observations to another. Further, in contrast to the matched-pairs experiment, the pair effects will not cancel when you calculate (Y 1 Y 2). Compare V(Y 1Y 2) = V(D)for this design with the matched-pairs design of Exercise 12.31. Why is the variance for the completely randomized design usually larger?1
Unfortunately, we don't have that question answered yet. But you can get it answered in just 5 hours by Logging in or Becoming a subscriber.
Becoming a subscriber
Or look for another answer