DSC04871

anu rui* »evcrai rngn ouuiers. i ric* nnoimrap i page /71* a eompiMeiMntemlv& iool for getting accuratc confidenee intcrval§ without the normality condition. Three bootstrap s/ rn u I ations, eacft wlth 10,000 repetitions, gxva these 95% confidenee intervals for mean sale# In the entfre region: (20.42, 27.26), (20.40, 27.18), and (20,48, 27.28).

(b) Find the 95% t confidenee intervai for the mean. It is essentially the sarnę a* the bootstrap lntervals. The lesson is that for sample s i/e s as largeasw * 50, / procedures are very robust.

7.10    What critical value t* from Table D should be used for a confidenee intenai

for the mean of the population in eaeh of the following situations?

(a)    A 95% confidenee intemil based on n ^m 15 observatlons.

(b)    A 90% confidenee interval from an SRS of 25 observations.

(c)    An 80% confidenee interval from a sample of size 30.

7.11    Use software to find the critical values t* that you would use for 95%

confidenee intcrvals for sample si/es of 10, 20, 30, 40, 50, 100, 200, asilOO.

Plot the values of /* versus the sample si/e and describe the relationship.

7.12    The one-samplc / Itatiftic for testing

Ho¹. M - 0 H_a\ H > 0

i ron i a sample of n

20 obiemtions has the value /    1.87.