Refer to Exercise 13.78. Suppose that a balanced

Chapter 13, Problem 79SE

(choose chapter or problem)

Problem 79SE

Refer to Exercise 13.78. Suppose that a balanced completely randomized design is to be employed and that prior experimentation suggests that σ = 20.

a How many replications would be required to estimate any treatment (drug combination) mean correct to within ±10 with probability .95?

b How many degrees of freedom will be available for estimating σ 2 when using the number of replications determined in part (a)?

c Give the approximate half-width of a 95% confidence interval for the difference in mean responses for two treatments when using the number of replications determined in part (a).

Reference

A study was initiated to investigate the effect of two drugs, administered simultaneously, on reducing human blood pressure. It was decided to use three levels of each drug and to include all nine combinations in the experiment. Nine high-blood-pressure patients were selected for the experiment, and one was randomly assigned to each of the nine drug combinations. The response observed was a drop in blood pressure over a fixed interval of time.

a Is this a randomized block design?

b Suppose that two patients were randomly assigned to each of the nine drug combinations. What type of experimental design is this?