Chapter 4.2, Problem 78E

(a)

To determine

To design an experiment that uses a completely randomized design to investigate this question.

Explanation of Solution

In this study, medical researchers want to compare the success rate of a new non-invasive method for replacing heart valves using a cardiac catheter with the traditional open heart surgery.

If we conduct a completely randomized experiment then we will randomly assign all the patients a unique numbers. Then we will choose the first randomly selected half of the patients to go through the new non-invasive method for replacing heart valves using cardiac catheter i.e. the twenty male patients will go through it. And the remaining patients will go through the traditional open heart surgery method. Then both of the groups will be compared on the basis of the success rate.

(b)

To determine

To design an experiment that uses a matched pairs design to investigate this question and explain your method of pairing.

Explanation of Solution

In this study, medical researchers want to compare the success rate of a new non-invasive method for replacing heart valves using a cardiac catheter with the traditional open heart surgery.

A matched pairs experiment compares two treatments which each subject undergoes. Thus, in this case we cannot do the matched pair experiment as the patient cannot go through both of the surgery in their life. It is therefore not possible for the patients to go through the both of the treatments and thus cannot be compared using this design of the experiment.

(c)

To determine

To explain which design do you prefer.

Answer to Problem 78E

Completely randomized experiment.

Explanation of Solution

The design of experiment we will choose be the completely randomized experiment as for matched pairs experiment, it compares two treatments which each subject undergoes. Thus, in this case we cannot do the matched pair experiment as the patient cannot go through both of the surgery in their life. It is therefore not possible for the patients to go through the both of the treatments and thus cannot be compared using this design of the experiment.