In: Statistics and Probability
For a binomial parameter, p, we would like to test H 0 : p 0 = 0.05 against H 1 : p 0 ≠ 0.05 at the 5% level of significance. Assume we have a sample of 200 trials.
a. Is 200 large enough to use our large-sample test for p? Answer Yes or No and present your evidence as a series of inequalities.
b. What test statistic would you use for the test?
c. What sampling distribution would you use for the test?
d. State the decision rule you would use to perform the test
e. What is your conclusion if your sample contains 16 successes out of the 200 trials? Explain your choice.
f. Would your conclusion change if we changed H 1 to p 0 > 0.05 ? Answer Yes or No and explain.