X1,X2,...,Xn ~ N(μ,σ^2) =>
Xbar = 1/nΣXi
χ^2
=
Σ(Xi-Xbar)^2
-------------------
σ^2
=
nS^2
----
σ^2
~ χ^2(n-1) (chi-squared distribution with n-1 degrees of freedom)
------------------------------------------------
X1,X2,...,Xn ~ N(μ,σ^2) =>
Xbar ~ N(μ,σ^2/n) =>
Z =
Xbar-μ
----------
σ/root(n)
~ N(0,1)
------------------------------------------------
X1,X2,...,Xm ~ N(μx,σx^2)
Y1,Y2,...,Yn ~ N(μy,σy^2) =>
Xbar ~ N(μx,σx^2/m)
Ybar ~ N(μy,σy^2/n) =>
Xbar-Ybar ~ N(μx-μy,σx^2/m+σy^2/n) =>
Z =
(Xbar-Ybar)-(μx-μy)
---------------------
root(σx^2/m+σy^2/n)
~ N(0,1)
------------------------------------------------
X1,X2,...,Xn ~ N(μ,σ^2) =>
T =
root(n-1)(Xbar-μ)
-----------------
S
~ t(n-1) (Student's t distribution with n-1 degrees of freedom)
used in test of μ
------------------------------------------------
X1,X2,...,Xm ~ N(μx,σx^2)
Y1,Y2,...,Yn ~ N(μy,σy^2)
σx^2 = σy^2
=>
T =
root(m+n-2){(Xbar-Ybar)-(μx-μy)}
--------------------------------------
root((1/m+1/n)(mSx^2+nSy^2))
~ t(m+n-2)
used in test of μx-μy
Write a comment