Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

Chapter 5 Decision Making For Two Samples

§5.1 Introduction

Population 1: N(µ1 , σ12 ) ⇐= X11 , X12 , . . . , X1n1

Population 2: N(µ2 , σ22 ) ⇐= X21 , X22 , . . . , X2n2

We are interested in comparing these two populations, such as

H0 : µ 1 = µ 2 vs H1 : µ1 6= µ2

or
H0 : σ12 = σ22 vs H1 : σ12 6= σ22

§5.2 Inference on the means of two populations, variances known

1. Assumptions:

(1) X11 , X12 , . . . , X1n1 are a random sample from population 1 ( );

(2) X21 , X22 , . . . , X2n2 are a random sample from population 2 ( );

(3) Two populations are ;

(4) Both populations are normally distributed with .

(Note: If not normal, CLT will apply for large n1 and n2 .)

2. Hypothesis testing on the difference in means: Two-sample Z-test

(a) H1 : µ1 − µ2 6= ∆0

(1) H0 : µ1 − µ2 = ∆0 vs (b) H1 : µ1 − µ2 > ∆0

(c) H1 : µ1 − µ2 < ∆0

Note that when ∆0 = 0, we often write as H0 : µ1 = µ2 vs H1 : µ1 6= µ2 etc

66
(2) Test statistic:
Z0 =

(3) Rejection region: Reject H0 if

(a)

(b)

(c)

Remark: p-values (a) = 2[1 − Φ(|z0 |)]

(b) = 1 − Φ(z0 )

(c) = Φ(z0 )

Example 5.1 (ex. 5.2 on page 223) Two types of plastic are suitable for use by an electronics
component manufacture. The breaking strength of this plastic is important. It is known
that σ1 = σ2 = 1.0 psi. From a random sample of size n1 = 10 and n2 = 12, we obtain
x̄1 = 162.7 and x̄2 = 155.4. The company will not adopt plastic 1 unless its mean breaking
strength exceeds that of plastic 2 by 10 psi. Based on the sample information, should it use
plastic 1? Assume that both populations are normal. Choose α = 0.05.

i) H0 : vs H1 :

ii) Test statistic:

iii) Rejection region:

iv) Conclusion:

Interpretation:

67
3. Confidence interval on the difference in means (variance known):

Since
X̄1 − X̄2 − (µ1 − µ2 )
Z= r
σ12 σ22
n1
+ n2

has a standard normal distribution N(0, 1), then we can get 100(1 − α)% confidence interval
for the mean difference µ1 − µ2 is

Similarly, we could get 100(1 − α)% LCB or UCB for the mean difference µ1 − µ2 .

Note that the above results can also be used to draw the conclusion for the corresponding
testing problem.

Example 5.2 Tensile strength tests were performed on two different grades of aluminum
spars used in manufacturing process and the testing procedure, the standard deviations of
tensile strength are assumed to be known. The data obtained are shown below. Find a 90%
CI on the difference in means. Assume two populations are normal and independent.

Spar Grade Sample size Sample mean(Tensile strength) Standard deviation

1 n1 = 10 x̄1 = 87.6 (kg/mm2 ) σ1 = 1.0(kg/mm2 )


2 n2 = 12 x̄2 = 74.5 (kg/mm2 ) σ2 = 1.5(kg/mm2 )

68

You might also like