The parameter
0.01, 0.05, 0.1, corresponding to confidence levels of 0.99, 0.95, and
0.90, respectively.
Confidence intervals for the population mean when the
population variance is known
Let X be the mean of a random sample of size n, drawn from an infinite
population with known standard deviation . The 100(1- ) % [i.e., 99%, 95%,
90%, etc.], central, two-sided confidence interval for the population mean
( X z
/ n , X+z
/2
is exceeded with a probability of
mean, X, is
/ n.
The one-sided upper and lower 100(1- ) % confidence limits for the population
mean
are, respectively, X+z
sided, confidence interval is defined as (- , X+z
sided, confidence interval as (X z
intervals we use the value z , rather than z
In general, the value z
value of z whose probability of exceedence is k, i.e., Pr[Z>z
= 1 – k. The normal distribution was described in Chapter 17.
Confidence intervals for the population mean when the
population variance is unknown
Let X and S, respectively, be the mean and standard deviation of a random
sample of size n, drawn from an infinite population that follows the normal
distribution with unknown standard deviation . The 100 (1
95%, 90%, etc.] central two-sided confidence interval for the population mean
is ( X t
n-1, /2
variate with
= n-1 degrees of freedom and probability /2 of exceedence.
The one-sided upper and lower 100 (1- ) % confidence limits for the
population mean
is known as the significance level. Typical values of
/ n ), where z
/2
/ n , and X z
in the standard normal distribution is defined as that
k
S / n , X+ t
n-1, /2
are, respectively,
X + t
S/ n , and X t
n-1, /2
is a standard normal variate that
/2
/2. The standard error of the sample
/ n . Thus, a lower, one-
/ n), and an upper, one-
/ n,+ ). Notice that in these last two
.
S/ n ), where t
n-1, /2
] = k, or Pr[Z<z
k
) % [i.e., 99%,
is Student's t
n-1, /2
S / n.
Page 18-24
are
is
]
k