单比例检验的精确方法的详细公式:
Hypotheses
The null hypothesis for a 1 proportion test is: H0: p = p0 where
· p = the population proportion
· p0 = the hypothesized proportion
You can choose any one of three alternative hypotheses:
H1: p > p0
One-tailed test
H1: p < p0
One-tailed test
H1: p p0
Two-tailed test
Confidence interval
Lower limit (PL) =
n1 * F
n2 + n1 * F
where
n1 = 2x
n2 = 2 (n - x +1)
x = number of events
n = number of trials
F = lower a/2 point of F with n1 and n2 degrees of freedom
Upper limit (PU) =
n1 * F
n2 + n1 * F
where
n1 = 2 (x + 1)
n2 = 2 (n - x)
x = number of events
n = number of trials
F = upper a / 2 point of F with n1 and n2 degrees of freedom
The interval (PL , PU) is an approximate 100 (1 - a)% confidence interval of p.
Note: When x = 0 or x = N, Minitab calculates only a one-sided confidence interval.
where:
= observed probability, = x / n, where x is the observed number of events in n trials
p0 = hypothesized probability
n = the number of trials
等等....Minitab帮助里面的“另请参见”里面有详细的公式啊 |
发表于 2011-11-6 05:02:25
收藏
楼主