Linear combinations of Normal random variables

Linear combinations of Normal random variables

Spec 1.1 · Edexcel International A Level Further Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • E(aX±bY)=aE(X)±bE(Y)\mathrm{E}(aX\pm bY)=a\mathrm{E}(X)\pm b\mathrm{E}(Y) always.
  • For independent variables, Var(aX±bY)=a2Var(X)+b2Var(Y)\mathrm{Var}(aX\pm bY)=a^2\mathrm{Var}(X)+b^2\mathrm{Var}(Y): variances always add.
  • A linear combination of independent Normals is Normal.
  • X1+⋯+Xn∼N(nμ,nσ2)X_1+\dots+X_n\sim\mathrm{N}(n\mu,n\sigma^2) but nX∼N(nμ,n2σ2)nX\sim\mathrm{N}(n\mu,n^2\sigma^2).

Means and variances

Means follow the signs; variances add and each coefficient is squared.

A linear combination of random variables is an expression such as aX+bYaX+bY or 3X−2Y3X-2Y. For any XX and YY the mean follows the signs. For independent XX and YY the variance adds, with each coefficient squared.

  • MeanE(aX±bY)=aE(X)±bE(Y)\mathrm{E}(aX\pm bY)=a\mathrm{E}(X)\pm b\mathrm{E}(Y)
  • Variancea2Var(X)+b2Var(Y)a^2\mathrm{Var}(X)+b^2\mathrm{Var}(Y)

XX and YY are independent with Var(X)=4\mathrm{Var}(X)=4 and Var(Y)=9\mathrm{Var}(Y)=9. What is Var(2X−Y)\mathrm{Var}(2X-Y)?

Normal variables

A linear combination of independent Normal variables is also Normal.

If X∼N(μx,σx2)X\sim\mathrm{N}(\mu_x,\sigma_x^2) and Y∼N(μy,σy2)Y\sim\mathrm{N}(\mu_y,\sigma_y^2) are independent, then any linear combination is Normal. No proof is required.

Write the new distribution with its mean and variance, then standardise with the standard deviation to find probabilities.

  • aX ± bY∼N(aμx±bμy, a2σx2+b2σy2)\sim\mathrm{N}(a\mu_x\pm b\mu_y,\ a^2\sigma_x^2+b^2\sigma_y^2)

X∼N(10,22)X\sim\mathrm{N}(10,2^2) and Y∼N(4,32)Y\sim\mathrm{N}(4,3^2) are independent. What is the distribution of X+YX+Y?

Sums and differences

For X−YX-Y the means subtract but the variances still add.

For a sum, X+Y∼N(μx+μy, σx2+σy2)X+Y\sim\mathrm{N}(\mu_x+\mu_y,\ \sigma_x^2+\sigma_y^2). For a difference, X−Y∼N(μx−μy, σx2+σy2)X-Y\sim\mathrm{N}(\mu_x-\mu_y,\ \sigma_x^2+\sigma_y^2).

Comparisons such as "is XX bigger than YY?" become P(X−Y>0)\mathrm{P}(X-Y>0).

−15−10−55101520250.020.040.060.080.1xy0N(5, 25)
X−Y∼N(5,25)X-Y\sim\mathrm{N}(5,25). The shaded area to the right of 00 is P(X>Y)=0.841\mathrm{P}(X>Y)=0.841.

Worked example

X∼N(20,42)X\sim\mathrm{N}(20,4^2) and Y∼N(15,32)Y\sim\mathrm{N}(15,3^2) are independent. Find P(X>Y)\mathrm{P}(X>Y).

X∼N(8,4)X\sim\mathrm{N}(8,4) and Y∼N(3,5)Y\sim\mathrm{N}(3,5) are independent. What is the variance of X−YX-Y?

Coefficients

Each coefficient is squared in the variance.

For X∼N(12,32)X\sim\mathrm{N}(12,3^2) and Y∼N(7,22)Y\sim\mathrm{N}(7,2^2), the mean of 3X−2Y3X-2Y is 3×12−2×7=223\times12-2\times7=22. The variance is 32×9+22×4=973^2\times9+2^2\times4=97, so 3X−2Y∼N(22,97)3X-2Y\sim\mathrm{N}(22,97).

  • Mean of 3X − 2Y3×12 − 2×7 = 22
  • Variance of 3X − 2Y3²×9 + 2²×4 = 97

Worked example

Find P(3X−2Y>30)\mathrm{P}(3X-2Y>30) for the variables above.

X∼N(5,4)X\sim\mathrm{N}(5,4). What is the variance of 5X5X?

Several observations

A total of nn independent observations has variance nσ2n\sigma^2, but nXnX has variance n2σ2n^2\sigma^2.

Let X1,…,XnX_1,\dots,X_n be independent observations from N(μ,σ2)\mathrm{N}(\mu,\sigma^2). Their total varies less than one observation multiplied by nn, because independent errors partly cancel.

Use X1+X2X_1+X_2 when two separate items are involved and 2X2X when the same item is counted twice.

X₁ + … + Xₙ

Mean:
nμ
Variance:
nσ²
Use when:
n separate items

nX

Mean:
nμ
Variance:
n²σ²
Use when:
one item counted n times

X∼N(10,3)X\sim\mathrm{N}(10,3). What is the variance of the total of three independent observations X1+X2+X3X_1+X_2+X_3?

Worked example

Define the new variable, write down its mean and variance, state the distribution, then standardise.

Passenger masses are N(75,122)\mathrm{N}(75,12^2) kg and luggage masses N(18,52)\mathrm{N}(18,5^2) kg, all independent. Four passengers each carry one bag.

  1. 1

    Define

    The new variable W

  2. 2

    Mean and variance

    Using the rules

  3. 3

    State the distribution

    W∼N(…)W\sim\mathrm{N}(\ldots)

  4. 4

    Standardise

    Then use tables

Method

Worked example

Find the probability that the total mass of four passengers and their four bags exceeds 400400 kg.

Why is the variance of WW equal to 4(144+25)4(144+25) and not 16(144+25)16(144+25)?

Try an exam question

X∼N(12,32)X\sim\mathrm{N}(12,3^2) and Y∼N(7,22)Y\sim\mathrm{N}(7,2^2) are independent. Find P(X+2Y>30)\mathrm{P}(X+2Y>30).

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

What is the distribution of the total time X+YX+Y?

1 mark, marked against the mark scheme

Revision notes