Salam Academy
Option · Module

Combinations of random variables

3 subtopics≈ 100 min in total
  1. E and Var of aX + b and aX + bY

    Expectations behave exactly as you would hope: E(aX+bY)=aE(X)+bE(Y)E(aX+bY)=aE(X)+bE(Y). Variances need care: constants are squared, added constants vanish, and for independent variables the variances always add, even for a difference.

    ≈ 30 min
  2. Sums of observations and multiples

    Weighing one bag and multiplying by 44 is not the same as weighing four different bags. Both totals have the same mean, but 4X4X has variance 16σ216\sigma^2 while X1+X2+X3+X4X_1+X_2+X_3+X_4 has only 4σ24\sigma^2.

    ≈ 35 min
  3. Combining normal and Poisson variables

    Any linear combination of independent normals is normal, so "is XX bigger than YY?" becomes "is X−YX-Y positive?". Poisson variables add to a Poisson, but only sums: 2X2X and X−YX-Y are not Poisson.

    ≈ 35 min