INTERVAL ESTIMATE

• If a large number of such intervals were computed, then of them will contain ? .

• Now, how do we obtain ? and ? ?

• For a large number of observations ( ), we can assume that the sample mean ( ) is a NORMALLY DISTRIBUTED random variable.

• This suggests that we can obtain the confidence interval from N(0,1), i.e.

• Rearranging the above equation we obtain:

where

N(0,1)

Z1

Z2

Previous slide

Next slide

Back to first slide

View graphic version