320 CHAPTER 6 Inferential Statistics
P(Y =y) =
∑∞
j=1
P(Y =y|X=xi)P(X=xi). (6.2)
Having noted this, we now proceed:
μX+Y =
∑∞
i=1
∑∞
j=1
(xi+yj)P(X=xi andY =yj)
=
∑∞
i=1
∑∞
j=1
xiP(X=xi andY =yj)
+
∑∞
i=1
∑∞
j=1
yjP(X=xiandY =yj)
=
∑∞
i=1
∑∞
j=1
xiP(X=xi|Y =yj)P(Y =yj)
+
∑∞
i=1
∑∞
j=1
yjP(Y =yj|X=xi)P(X=xi)
=
∑∞
i=1
xi
∑∞
j=1
P(X=xi|Y =yj)P(Y =yj)
+
∑∞
j=1
yj
∑∞
i=1
P(Y =yj|X=xi)P(X=xi)
=
∑∞
i=1
xiP(X=xi) +
∑∞
j=1
yiP(Y =yi) by (6.1) and (6.2)
= μX+μY,
proving that
E(X+Y) = E(X) +E(Y). (6.3)
Next, note that ifXis a random variable and ifaandbare constants,
then it’s clear that E(aX) =aE(X); from this we immediately infer
(sincebcan be regarded itself as a random variable with meanb) that
E(aX+b) = aE(X) +b.