Академический Документы
Профессиональный Документы
Культура Документы
Lower
Confidence
Limit
Upper
Point Estimate
Width of
confidence interval
Confidence
Limit
x X Z
(where Z is the normal distributions critical valuenfor a probability of /2 in each
tail)
.05
.475
.025
2
Z= -1.96
Lower
Confidence
Limit
/ 2 .025
.475
0
Point
Point Estimate
Estimate
.025
2
Z= 1.96
Upper
Confidence
Limit
Example:
Suppose there are 69 U.S. and imported beer brands in the
U.S. market. We have collected 2 different samples of 25
brands and gathered information about the price of a 6-pack,
the calories, and the percent of alcohol content for each
brand. Further, suppose that we know the population
standard deviation ( ) of price is $1.45. Here are the
samples information:
Sample A: Mean=$5.20, Std.Dev.=$1.41=S
Sample B: Mean=$5.59, Std.Dev.=$1.27=S
From sample B
We are 95% confident that the true mean price is between $5.02
and $6.16. (Failed)
We are 99% confident that the true mean price is between $4.84
and $6.36.
Students t distribution
Note: t
Z as n increases
See our beer example
Standard
Normal
t (df = 13)
t-distributions are bell-shaped
and symmetric, but have
fatter tails than the normal
t (df = 5)
Using
Z
( X )
X Z*
n Z
e
2
Sampling Error, e
2
To determine the required sample size for the mean, you must know:
1.
1.
2.
2.
3.