Mean and Variance of the F-distribution
📂Probability DistributionMean and Variance of the F-distribution
X∼F(r1,r2) Surface Area
E(X)=r2−2r2,r2>2Var(X)=d1(d2−2)2(d2−4)2d22(d1+d2−2),r2>4
Derivation
Strategy: Like the chi-squared distribution, the F-distribution also has known moment-generating functions, which we will use.
Moments of the F-Distribution: Let’s say X∼F(r1,r2) and it can be expressed as X=X2X1. If X1 and X2 both follow chi-squared distributions with degrees of freedom d1,d2 and if d2>2k then the k-th moment exists
EXk=(r1r2)kEX1kEX2−k
Moments of the Chi-Squared Distribution: Let’s say X∼χ2(r). If k>−r/2 then the k-th moment exists
EXk=Γ(r/2)2kΓ(r/2+k)
Mean
Assuming r2>2 then because −k=−1>−r2/2 thus, EX2−1 exists.
If k=1 then according to the moment-generating functions
EX1====(r1r2)1EX11EX2−1r1r2Γ(r1/2)21Γ(r1/2+1)Γ(r2/2)2−1Γ(r2/2−1)r1r22r1r2/2−11r2−2r2
■
Variance
Assuming r2>4 then because −k=−2>−r2/2 thus, EX2−2 exists.
If k=2 then according to the moment-generating functions
EX2=====(r1r2)2EX12EX2−2(r1r2)2Γ(r1/2)22Γ(r1/2+2)Γ(r2/2)2−2Γ(r2/2−2)(r1r2)2(r2/2−1)(r2/2−2)(r1/2+1)r1/2(r1r2)2(r2−2)(r2−4)(r1+2)r1r1(r2−2)(r2−4)r22(r1+2)
Therefore
Var(X)===r1(r2−2)(r2−4)r22(r1+2)−[r2−2r2]2r1(r2−2)2(r2−4)r22[(r1+2)(r2−2)−r1(r2−4)]r1(r2−2)2(r2−4)2r22(r1+r2−2)
■