Question

Suppose you want to test the health of panda bears in a local region of a country and you want to measure their health by the amount of food, in pounds, for which they eat in a controlled setting in a day. You do this study for two different regions in that country and obtain the following results:

S1={23,32,19,21,18,14,32,42,39,27,29,24,18}

S2={32,19,25,18,12,15,27,23,19,21,18,17,29,30,32}

From other research studies, you determine that some panda bears eat more than others, but all healthy panda bears tend to eat more than 20 pounds of food in any given day.

a)For population 1, what is a point estimate for the proportion of panda bears that eat a healthy amount of food in any given day? Round to 4 decimal places.

b)For population 2, what is a point estimate for the proportion of panda bears that eat a healthy amount of food in any given day? Round to 4 decimal places.

c)What is the standard error associated to the estimate in the difference in proportions between the two populations of interest?

d)What is the marginal error for a two-sided confidence interval for the population mean difference such that the probability of committing a Type I error is 3%? Round to 4 decimal places.

e)

Assume we define the population mean difference point estimate in the direction:

pd:=pa-pb

such that the value of pd is positive. Consider the null and alternative hypothesis pair:

H0 |pd=0

Ha |pd doesn't = 0

In continuation of the previous problems, we then have: (Round both boundaries to 2 decimal places)

Answer 1