You are on page 1of 10

Inferences on Two Populations

A. Inferences on Two Proportions


Sample Problem
Two different types of injection-molding machines are used to form plastic parts. A part is
considered defective if it has excessive shrinkage or is discolored. Two random samples, each of
size 300, are selected, and 15 defective parts are found in the sample from machine 1, and 8
defective parts are found in the sample from machine 2.

a) Is it reasonable to conclude that both machines produce the same fraction of defective parts,
using α = 0.05?

Let p1= proportion of defective parts for Machine 1


P2= proportion of defective parts for Machine 2

x 1=15
x 2=8
n1=n2=300
Verify the conditions required.
1. The sample proportions are from two simple random samples.
2. The two samples are independent.
3. There must be at least 5 successes and 5 failures from each of the two samples (
n1 p1 ≥ 5,n1 (1−p 1)≥5 , n2 p2 ≥ 5, and n2 (1−p 2)≥5

Step 1: Determine the null and alternative hypothesis.

H o : p1= p2∨( p1− p2=0)

H A : p1 ≠ p2 ∨( p1− p 2 ≠ 0)
Step 2. Determine the rejection region.

Since the parameter of interest is the mean difference between two population
proportions, we use the z-table.

For a two-tailed, 0.05 significance level,

z 0.05=± 1.960

Rejection rule: if the test statistic is less than -1.960 or greater than +1.960, reject the
null hypothesis.

Step 3. Compute the test Statistic

Solve for pooled proportion

ENS185 Ruda-Bayor
x 1+ x2
p=
n1 + n2
15+8
p= =0.0383
300+300
Solve for the test statistic:

( p ¿ ¿ 1−p 2)
z=( ^ p 2 )−
p1 −^ ¿

√ p ( 1− p )
(1 1
+
n1 n2 )
z=
( 300 300 )
15

8
−(0)
=1.49

√0.0383 (1−0.0383)( 300 + 300 )


1 1

Step 4. Make the decision


Since the test statistic is not within the rejection region, we fail to reject the null hypothesis.

Step 5. State your conclusion.


At 0.05 significance level, there is not enough evidence to conclude that both machines produce
unequal fraction of defective parts. Practically, Machine 1 and 2 produce is observed to produce
the same proportion of defective parts.

a) Create the appropriate confidence interval.

A 95% confidence interval can be used to supplement a two-tailed 0.05 significance level
hypothesis is test.

Step 1. Compute for the point estimate.

The point estimate for the difference of two population proportions is equal to ^
p1 − ^
p2 .
From the problem above,

15 8
^
p1 − ^
p2= − =0.0233
300 300
Step 2. Determine the critical value.

For a 95% confidence level,

z 0.05=1.960

Step 3. Compute the margin of error.

ENS185 Ruda-Bayor
E=z α ×
2 √ p 1(1−^
^
n1
p1 ) ^
+
p 2(1−^
n2
p2 )


15 15 8 8
(1− ) (1− )
300 300 300 300
E=1.960× + =0.0307
300 300
Step 4. Compute the confidence interval.

Lower Limit

(^ p 2) −E=0.0233−0.0307=−0.0073
p1− ^

Upper Limit

(^ p 2) + E=0.0233+ 0.0307=0.0540
p1− ^

Step 5. Make a conclusion.

With 95% confidence, the true difference between the population proportions of defects
between machine 1 and 2 is from -0.0073 and 0.0540.

Since 0 is contained within the interval, we cannot conclude that there is a significant
difference between the two population proportions using the 95% confidence level.

B. Inferences on Two Population Means


Case 1: Known Population Variances
Sample Problem
Two types of plastic are suitable for an electronics component manufacturer to use. The
breaking strength of this plastic is important. It is known that σ 1=σ 2=1.0 psi. From a random
sample size of n1=10 and n2=12, you obtain x 1=162.5 and x 2=155.0. The company will not
adopt plastic 1 unless its mean breaking strength exceeds that of plastic 2 by at least 10 psi.
Assume that the populations follow normal distribution.

a) Based on the information, should it use plastic 1? Use 0.01 significance level.

Verify the conditions required.


1. The values of σ1 and σ2 are known
2. The two samples are independent.
3. Both samples are simple random samples.
4. Either or both of these conditions are satisfied: The two sample sizes are both large
(with n1 > 30 and n2 > 30) or both samples come from populations having normal
distributions.

ENS185 Ruda-Bayor
Step 1: Determine the null and alternative hypothesis.

H o : μ1−μ2 ≥10

H A : μ1−μ 2<10

Step 2. Determine the rejection region.

Since the parameter of interest is the mean difference between two population means
with known variances, we use the z-table.

For a left-tailed, 0.01 significance level,

z 0.01=−2.323

Rejection rule: if the test statistic is less than -2.323, reject the null hypothesis.

Step 3. Compute the test Statistic

Solve for the test statistic:

( x 1−x 2 )−(μ 1−μ2 )


z=


2 2
σ1 σ 1
+
n 1 n2

( 162.5−155 )−(10)
z= =−5.839

√ 12 12
+
10 12
Step 4. Make the decision
Since the test statistic is within the rejection region, we reject the null hypothesis.

Step 5. State your conclusion.


At 0.05 significance level, there is enough evidence to conclude that difference between the
mean breaking point of Plastic 1 and mean breaking point of Plastic 2 is less than 10 psi.
Practically, this means the company should not adopt plastic 1 since the mean breaking strength
of plastic does not seem to exceed the mean breaking strength of Plastic 2 by at least 10 psi.

b) Construct the 98% confidence interval.

A 98% confidence interval can be used to supplement a one-tailed 0.01 significance level
hypothesis is test.

Step 1. Compute for the point estimate.

The point estimate for the difference of two population proportions is equal to x 1−x 2

From the problem above,

ENS185 Ruda-Bayor
x 1−x 2=162.5−155.0=7.5
Step 2. Determine the critical value.

For a 98% confidence level,

z 0.02=2.323

Step 3. Compute the margin of error.

E E=z α
2 √ σ 21 σ 22
+
n1 n 2

E=2.323
√ 12 12
+ =0.9961
10 12

Step 4. Compute the confidence interval.

Lower Limit

( x 1−x 2 ) ± E=7.5−0.9961=6.5049
Upper Limit

( x 1−x 2 ) ± E=7.5+0.9961=8.4960

Step 5. Make a conclusion.

With 98% confidence, the mean breaking point of Plastic 1 is higher than the mean
breaking point of Plastic 2 by 6.5049 psi to 8.4960 psi.

Since both the lower and upper limits do not exceed at least 10 psi, it is not advisable to
adopt Plastic 1 based on the set requirements.

Case 2: Unknown Population Variances, Equal


Sample Problem
Two Catalysts are being analyzed to determine how they affect the mean yield of a chemical
process. Specifically, catalyst 1 is currently used; but catalyst 2 is acceptable. Because catalyst 2 is
cheaper, it should be adopted, if it does not change the process yield. A test run on 8 samples
each yielded the following results: x 1=92.26 , x 2=92.733 , s1=2.39 , s2=2.98 . Is there any
difference in the mean yields? Assume equal variances and normal distribution. Use 0.10
significance level.

Verify the conditions required.


1. The values of σ1 and σ2 are unknown
2. The two samples are independent.
3. Both samples are simple random samples.

ENS185 Ruda-Bayor
4. Either or both of these conditions are satisfied: The two sample sizes are both large
(with n1 > 30 and n2 > 30) or both samples come from populations having normal
distributions.

Step 1: Determine the null and alternative hypothesis.

H o : μ1−μ2=0

H A : μ1−μ 2 ≠ 0

Step 2. Determine the rejection region.

Since the parameter of interest is the mean difference between two population means
with unknown population variances, we use the t-table.

Solve for the degrees of freedom:

df =n1 +n2−2

df =8+8−2=14
For a two-tailed, 0.10 significance level,

t 0.10=± 1.761

Rejection rule: if the test statistic is less than -1.761 or greater than 1.761, reject the null
hypothesis. .

Step 3. Compute the test Statistic

Solve for the pooled variance

( n1−1 ) s 21+ ( n2−1 ) s22


2
s= p
( n1−1 ) +(n2−1)
( 8−1 ) 2.392+ ( 8−1 ) 2.982
s2p= =7.2963
( 8−1 ) +(8−1)
Solve for the test statistic:

( x 1−x 2) −(μ 1−μ2)


t=


2 2
sp sp
+
n1 n2

( 92.255−92.733 )−(0)
t= =−0.354

√ 7.2963 7.2963
8
+
8

ENS185 Ruda-Bayor
Step 4. Make the decision
Since the test statistic is not less than -1.761, we fail to reject the null hypothesis.

Step 5. State your conclusion.


At 0.10 significance level, there is not enough evidence to conclude that the mean yield of the
process is different for Catalyst 1 and Catalyst 2.

Practical Conclusion: Since there seems to be no significant difference in the mean yield of the
process when using Catalyst 1 and Catalyst 2, Catalyst 2 is acceptable to use because it is cheaper.

Case 3: Unknown Population Variances, Unequal


Use the example above but this time, assume unequal variances.

Two Catalysts are being analyzed to determine how they affect the mean yield of a chemical
process. Specifically, catalyst 1 is currently used; but catalyst 2 is acceptable. Because catalyst 2 is
cheaper, it should be adopted, if it does not change the process yield. A test run on 8 samples
each yielded the following results: x 1=92.26 , x 2=92.733 , s1=2.39 , s2=2.98 . Is there any
difference in the mean yields? Assume unequal variances and normal distribution. Use 0.10
significance level.

Verify the conditions required.


1. The values of σ1 and σ2 are unknown
2. The two samples are independent.
3. Both samples are simple random samples.
4. Either or both of these conditions are satisfied: The two sample sizes are both large
(with n1 > 30 and n2 > 30) or both samples come from populations having normal
distributions.

Step 1: Determine the null and alternative hypothesis.

H o : μ1−μ2=0

H A : μ1−μ 2 ≠ 0

Step 2. Determine the rejection region.

Since the parameter of interest is the mean difference between two population means
with unknown population variances, we use the t-table.

Solve for the degrees of freedom:

ENS185 Ruda-Bayor
( )
2 2 2
s1 s2
+
n1 n 2
df =

( ) ( )
2 2 2 2
s1 s1
n1 n1
+
n1−1 n2−1

( )
2 2 2
2.39 2.98
+
8 8
df = =13.37 ≈ 13
( ) ( )
2 2 2 2
2.39 2.98
8 8
+
8−1 8−1

For a two-tailed, 0.10 significance level,

t 0.10=± 1.771

Rejection rule: if the test statistic is less than -1.771 or greater than 1.771, reject the null
hypothesis. .

Step 3. Compute the test Statistic

Solve for the test statistic:

( x 1−x 2) −( μ1−μ 2)
t=


2 2
s 1 s2
+
n 1 n2

( 92.255−92.733 )−(0)
t= =−0.354

√ 2.392 2.982
8
+
8

Step 4. Make the decision


Since the test statistic is not less than -1.761, we fail to reject the null hypothesis.

Step 5. State your conclusion.


At 0.10 significance level, there is not enough evidence to conclude that the mean yield of the
process is different for Catalyst 1 and Catalyst 2.

ENS185 Ruda-Bayor
Practical Conclusion: Since there seems to be no significant difference in the mean yield of the
process when using Catalyst 1 and Catalyst 2, Catalyst 2 is acceptable to use because it is cheaper.

C. Inferences on Two Standard Deviations


Sample Problem
Oxide layers on semiconductor wafers are etched in a mixture of gases to achieve the
proper thickness. The variability in the thickness of these oxide layers is a critical characteristic of
the wafer, and low variability is desirable for subsequent processing steps. Two different
mixtures of gases (M1 and M2) are being studied to determine whether one is superior in
reducing the variability of the oxide thickness. The results are as follows:

M1 M2
s1=1.54 s2=2.13
n1=9 n2 =7

Is there any evidence to indicate that the there is a significant difference in the two
variances? Use a fixed-level test with α = 0.05.

Step 1: Determine the null and alternative hypothesis.


2 2
H o :σ 1 =σ 2
2 2
H A :σ 1 ≠ σ 2
Note: More consistent means less dispersion.

This is a two-tailed test.

Step 2. Determine the rejection region.

Since the parameter of interest is the ratio between two variances, we use the F-table.

Since M2 has the greater variation (bigger standard deviation), we set it as the
numerator.

d f numerator =7−1=6

d f denominator=9−1=8

Since this is a two-tailed test, find the right critical value for α/2=0.05/2=0.025

F o > F1−α , n −1 , n −1
1 2

F 0.025 ,6 ,8 =4.65

ENS185 Ruda-Bayor
Again, we consider the right tailed critical value since we set that the numerator is always the
larger variation.

Rejection rule: if the test statistic is greater than 4.65, we reject the null hypothesis.

Step 3. Compute the test Statistic

Solve for the test statistic:


2
s numerator
F= 2
s denominator
2
2.13
F= 2
=1.913
1.54

Step 4. Make the decision


Since the test statistic is less than the critical value, we fail to reject the null hypothesis.

Step 5. State your conclusion.

At 0.05 significance level, there is not enough evidence to conclude that Mixture 1 has
significantly different variance compared to mixture 2.

Practical Conclusion: Since Mixture 1 produces significantly less variability compared to Mixture
2, the use of Mixture 1 is acceptable.

ENS185 Ruda-Bayor

You might also like