26. Housing, part 2. In Exercise 24, we found a 95% confidence interval to estimate the loss in home values.

a) Suppose the standard deviation of the losses was $3000 instead of the $1500 used for that interval. What would the larger standard deviation do to the width of the confidence interval (assuming the same level of confidence)?

b) Your classmate suggests that the margin of error in the interval could be reduced if the confidence level were changed to 90% instead of 95%. Do you agree with this statement? Why or why not?

c) Instead of changing the level of confidence, would it be more statistically appropriate to draw a bigger sample?