分位点
设随机变量X的分布函数为 F ( x ) F(x) F(x),若 x α x_\alpha xα使 P { X ≤ x α } = F ( x α ) = α P\{X\leq x_\alpha\}=F(x_\alpha)=\alpha P{X≤xα}=F(xα)=α,称 x α x_\alpha xα为此概率分布的(下侧) α \alpha α分位点
当 X X X~ N ( 0 , 1 ) N(0,1) N(0,1),分位点记为 u α u_\alpha uα或 z α z_\alpha zα
当 X X X~ χ 2 ( n ) \chi^2(n) χ2(n),分位点记为 χ α 2 ( n ) \chi_\alpha^2(n) χα2(n)
当 X X X~ t ( n ) t(n) t(n),分位点记为 t α ( n ) t_\alpha(n) tα(n)
当 X X X~ F ( m , n ) F(m,n) F(m,n),分位点记为 F α ( m , n ) F_\alpha(m,n) Fα(m,n)
性质
(1) − z α = z 1 − α , − t α ( n ) = t 1 − α ( n ) -z_\alpha = z_{1-\alpha},-t_\alpha(n) = t_{1-\alpha}(n) −zα=z1−α,−tα(n)=t1−α(n)
(2)
F
α
(
m
,
n
)
=
1
/
F
1
−
α
(
n
,
m
)
F_\alpha(m,n)=1/F_{1-\alpha}(n,m)
Fα(m,n)=1/F1−α(n,m)
证明:
P
{
X
≤
F
α
(
m
,
n
)
}
⇔
P
{
1
X
≥
1
F
α
(
m
,
n
)
}
⇔
1
−
P
{
1
X
≤
1
F
α
(
m
,
n
)
}
=
α
P\{X\leq F_\alpha(m,n)\}\Leftrightarrow P\{\frac 1X\geq \frac{1}{F_\alpha(m,n)}\}\Leftrightarrow 1-P\{\frac 1X\leq \frac{1}{F_\alpha(m,n)}\}=\alpha
P{X≤Fα(m,n)}⇔P{X1≥Fα(m,n)1}⇔1−P{X1≤Fα(m,n)1}=α
P { 1 X ≤ 1 F α ( m , n ) } = 1 − α P\{\frac 1X\leq \frac{1}{F_\alpha(m,n)}\}=1-\alpha P{X1≤Fα(m,n)1}=1−α相当于 P { x α ≤ F 1 − α ( n , m ) } = 1 − α P\{x_\alpha\leq{F_{1-\alpha}(n,m)}\}=1-\alpha P{xα≤F1−α(n,m)}=1−α,然后就可以证明了
渐进分布
设总体 X X X的分布函数为 F ( x ) F(x) F(x), E ( X ) = μ n , D ( X ) = σ F 2 E(X)=\mu_n,D(X)=\sigma_F^2 E(X)=μn,D(X)=σF2
则样本的均值的渐进分布为 N ( μ F , σ F 2 n ) N(\mu_F,\frac{\sigma_F^2}{n}) N(μF,nσF2)
n X n ˉ − μ F S n \sqrt n\frac{\bar{X_n}-\mu_F}{S_n} nSnXnˉ−μF近似服从 N ( 0 , 1 ) N(0,1) N(0,1)
设 ( X 1 , X 2 , . . . , X m ) , ( Y 1 , Y 2 , . . . , Y n ) (X_1,X_2,...,X_m),(Y_1,Y_2,...,Y_n) (X1,X2,...,Xm),(Y1,Y2,...,Yn)是来自总体 N ( μ 1 , σ 2 ) , N ( μ 2 , σ 2 ) , N(\mu_1,\sigma^2),N(\mu_2,\sigma^2), N(μ1,σ2),N(μ2,σ2),的样本,当m,n趋于无穷时
则 T = X ˉ − Y ˉ − ( μ 1 − μ 2 ) S 1 2 / m + S 2 2 / n T =\frac{\bar X - \bar Y - (\mu_1-\mu_2)}{\sqrt{S_1^2/m+S_2^2/n}} T=S12/m+S22/nXˉ−Yˉ−(μ1−μ2)近似服从 N ( 0 , 1 ) N(0,1) N(0,1)