STAT 211

Handout 9 (Chapter 9): Inferences Based on Two Samples

TESTS
1. One-sided (One tailed) test:
A. Lower tailed: H0: population characteristics  claimed constant value
(Left-sided) Ha: population characteristics < claimed constant value
B. Upper tailed: H0: population characteristics  claimed constant value
(Right-sided) Ha: population characteristics > claimed constant value
2. Two-sided (Two tailed) test: H0: population characteristics = claimed constant value
Ha: population characteristics  claimed constant value

Hypothesis testing for:

A. Independent Samples
I. Population characteristics: Difference between two population means,  1- 2.
0 is the claimed constant.
1 and  2 are the population means for X's and Y's, respectively.
_ _
x and y are the sample means for X's and Y's, respectively.
m and n are the sample sizes for X's and Y's, respectively.
 12 and  22 are the population variances for X's and Y's, respectively.
s12 and s 22 are the sample variances for X's and Y's, respectively.

Test statistics:
_ _
 x y    0
 z 
when both popn. distributions are normal and  1 ,  2 are known
2 2
 1  22
2

m n
 _ _

 x y    0
z  
when there is large sample size (m>40 and n>40) and  1 ,  2 are unknown
2 2
 2
s1 s 22

m n
 _ _

 x y    0
t  
 2 when both popn. distributions are normal and at least one sample size is
s1 s 22

m n
small where unknown  12 ,  22 are assumed to be different (  12   22 ) and degrees of
 2
 s12 s 22 
  
freedom, v  m n is used to look up the critical values. Your textbook calls this
s12 / m
2

 
s 22 / n
2

m 1 n 1
two-sample t test.
_ _
 x y    0
t  (m  1) s12  (n  1) s 22
 , s 2p  when both popn. distributions are normal and at
sp
1 1
 mn2
m n
least one sample size is small where unknown  12 ,  22 are assumed to be the same (  12 =  22
) and degrees of freedom, v  m  n  2 is used to look up the critical values. Your textbook
calls this pooled t test.

Decision can be made in one of the two ways:

a. Let z* or t* be the computed test statistic values.
if test statistics is z if test statistics is t
Lower tailed test P-value = P(z<z*) P-value = P(t<t*)
Upper tailed test P-value = P(z>z*) P-value = P(t>t*)
Two-tailed test P-value = 2P(z>|z*|)= 2P(z<-|z*|) P-value = 2P(t > |t*| )= 2P(t <- |t*| )

In each case, you can reject H0 if P-value   and fail to reject H0 (accept H0) if P-value > 

b. Rejection region for level  test:

if test statistics is z if test statistics is t
Lower tailed test z  -z t  -t;v
Upper tailed test z  z t  t;v
Two- tailed test z  -z/2 or z  z/2 t  -t/2;v or t  t/2;v

100(1-)% confidence Intervals with the same assumptions,

_ _  12  22
 x  y   z / 2 
  m n
_ _ s2 s2
 x  y   z / 2 1  2
  m n

_ _ s12 s 22
 x  y   t / 2 ; v  when you assume  12   22
  m n
_ _ 1 1
 x  y   t / 2 ; v s p  when you assume  12 =  22
  m n
Example: Two types of plastic are suitable for use by an electronics component manufacturer.
The breaking strength of this plastic is very important. It is known that  1 =  2 =1 psi. From a
 _ _
random sample of size n1 =10 and n 2 =12, we obtain x 1 =162.5 and x 2 =155. The company
will not adopt plastic 1 unless its mean breaking strength exceeds that of plastic 2 by at least 10
psi. Based on the sample information, would they use plastic 1? Use the significance level 0.05
in reaching a decision.

Exercise 9-2:
Sample sizes are large while population variances are unknown. Notice that this is a two-tailed
test.
_ _
 x y    0
  (42500  40400)  0
Test statistics: z   =4.8462
2 2
s1 s2 2200 2 1900 2
 
m n 45 45
(i) Reject H0 if z  -z/2=-1.96 or z  z/2 =1.96. z =4.8462 > 1.96 then reject H0. Conclude
that true average tread lives for two competing brand tires are different.
(ii) P-value = 2P(Z>4.8462)=2(0)=0. Since the P-value  =0.05, reject H0. Conclude that true
average tread lives for two competing brand tires are different.

If you prefer to answer the question computing the confidence interval, 95% confidence interval
s2 s2 2200 2 1900 2
would become  x  y   z / 2 1  2  (42500  40400)  1.96
_ _

  m n 45 45
=(1250.67, 2949.33). You would see that zero does not fall into interval and you would reject
H0.

Exercise 9-8:
Sample sizes are large while population variances are unknown. . Notice that this is a upper-
tailed test.
H 0 :  2  1  10 versus H a :  2  1  10
_ _
 y x   0
  (123.6  107.6)  10
Test statistics: z   =28.57
2 2
s1 s 1.3 2 22
 2 
m n 129 129

(i) Reject H0 if z  z =1.645 if =0.05. z =28.57 > 1.645 and reject H0. Conclude that the
data provide compelling evidence that the true average strength for the 1078 grade exceeds that
for the 1064 grade by more than 10.
(ii) P-value = P(Z>28.57)=0. Since the P-value  =0.05, reject H0. Conclude that the data
provide compelling evidence that the true average strength for the 1078 grade exceeds that for
the 1064 grade by more than 10.

Exercise 9-18:
Sample sizes are small and population variances are unknown. . Notice that this is a two-tailed
test.
 _ _
 x y    0
  ( 22.73  21.95)  0
Test statistics: t   =6.166
2 2
s1 s2 0.164 2 0.24 2
 
m n 6 5
(i) Reject H0 if t  -t/2;v= -2.447 or t  t/2;v =2.447 where =0.05 and v=6.886. t =6.166 >
2.447 and reject H0. Conclude that true average densities for two different types of brick are
different.
(ii) P-value = 2P(t>6.166)=2(0)=0. Since the P-value  =0.05, reject H0. Conclude that true
average densities for two different types of brick are different.

II. Population characteristics: Difference between two population proportions, p1-p2.

p0 is the claimed constant.
^ ^
p1 and p 2 are the sample proportions for X's and Y's, respectively.

p1 and p 2 are the population proportions for X's and Y's, respectively.
m and n are the large sample sizes for X's and Y's, respectively.

^ X Y m ^ n ^
The estimator for p is
p   p1  p 2
mn mn mn
 ^^
 p1  p 2   p0
z  
Test statistics:

  
^ ^11
p1 p  
 m n
Decision can be made in one of the two ways:

(a) Let z* or t* be the computed test statistic values.

Lower tailed test P-value = P(z<z*)
Upper tailed test P-value = P(z>z*)
 Two-tailed test P-value = 2P(z > |z*| )=2P(z <- |z*| )
In each case, you can reject H0 if P-value   and fail to reject H0 (accept H0) if P-value > 

(b) Rejection region for level  test:

Lower tailed test z  -z
Upper tailed test z  z
Two- tailed test z  -z/2 or z  z/2

100(1-)% large sample confidence Interval:


^^^^
^^ 11pp 1 21pp 2
  
p1 2zp /2 
 m n
Example: 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 from machine 1 while 8 defective
parts are found in the sample from machine 2. Is it reasonable to conclude that both machines
produce the same fraction of defective parts, using the significance 0.05? Find also the P-value
after answering the question.

Exercise 9-48(a):
Sample sizes are large enough to satisfy the assumptions and it is a two-tailed test.
^ X  Y 63  75 ^ 63 ^ 75
p  =0.2875 where p1  =0.21 and p 2  =0.4167
m  n 300  180 300 180

^^
p1 2pp 0
z   (0.21 0.4167) 0
Test statistics: =-4.844

^^ 1 0.2875(0. 125)( /30 1/ 80)

  
p 1     
  m n 
(i) Reject H0 if z  -z/2=-1.96 or z  z/2 =1.96. |z| =4.844 > 1.96 and reject H0. Conclude that
it is different for two groups of residents.
(ii) P-value = 2P(Z>4.844)=2(0)=0. Since the P-value  =0.05, reject H0. Conclude that it is
different for two groups of residents.

III.Population characteristics: Ratio of the two population variances,  12 /  22 or

standard deviations,  1 /  2 .
X and Y's are random sample from a normal distribution.
 12 and  22 are the population variances for X's and Y's, respectively.
s12 and s 22 are the sample variances for X's and Y's, respectively.
m and n are the sample sizes for X's and Y's, respectively.
s12
Test statistics: F  2
s2
Decision can be made in one of the two ways:

(a) Let F* be the computed test statistic values.

Lower tailed test P-value = P(F<F*)
Upper tailed test P-value = P(F>F*)
Two-tailed test P-value = 2P(F > F*)

In each case, you can reject H0 if P-value   and fail to reject H0 (accept H0) if P-value > 

(b) Rejection region for level  test:

 Lower tailed test F  F1-;m-1,n-1
Upper tailed test F  F;m-1,n-1
Two- tailed test F  F1-/2;m-1,n-1 or F  F/2;m-1,n-1
Notice that F1-/2;m-1,n-1 = 1 / F/2;n-1,m-1
100(1-)% confidence Interval for  1 /  2 :
2 2

s12 / s 22  12 s12 / s 22
 
F / 2;m 1,n 1  22 F1 / 2;m 1,n 1

Exercise 9.57:
(a) On the F-table, column for 5 and row for 8 will give the area on the right 0.05 with F0.05:5,8 =
3.69
(d) F0.95:8,5 = 1/ F0.05:5,8 = 1/3.69 =0.271
(e) The percentile is the area on the left of the value and it means the area on the right of the
value is 0.01. On the F-table, look at column for 10 and row for 12 with the area on the right
0.01.
P( F  F0.01:10,12 )= 0.99 then F0.01:10,12 = 4.31
(h) P(0.177  F  4.74) = P(F0.99:10,5  F  F0.05:10,5) =1-(0.01+0.05)=0.94 where F0.99:10,5 =
1/F0.01:5,10 = 1/5.64=0.177

Example: A study was performed to determine whether men and women differ in their
repeatability in assembling components on printed circuit boards. Two samples of 25 men and
21 women were selected and each subject assembled the units. The two sample standard
deviations of assembly time were smen=0.98 min and swomen=1.02 min. Is there evidence to
support the claim that men and women differ in repeatability for this assembly task? Use the
significance level 0.02 and state any necessary assumptions about underlying distribution of the
data.

Exercise 9.62:
 12
H 0 : 2  1 when H 0 :  12   22
2
 12
Ha :  1 when H a :  12   22
2 2

=0.10, m=48, n=45

s12 21.45 2
Test statistics: F  2   1.2219
s 2 19.45 2
Reject H0 if F  F1-/2;m-1,n-1 = F0.95;47,44 =1/1.6336=0.61214 or F  F/2;m-1,n-1= F0.05;47,44 =1.6412
Since the table on your book is not very detailed, you can use both degrees of freedom as 40 to
look at the table.
Since 1.2219 is not larger than 1.6412 or smaller than 0.61214, fail to reject H0 . Conclude that
the variances are the same.

B. Dependent Samples- Paired Data

Population characteristics: Difference between two population means, D =1-2.
 0 is the claimed constant.
Assumption: the difference distribution should be normal.
_
d  0 _
Test statistics: t  where D=X-Y and d and s D are the corresponding sample
sD / n
average and the standard deviation of D. Both X and Y must have n observations.

Decision can be made in one of the two ways:

(a) Let z* or t* be the computed test statistic values.

Lower tailed test P-value = P(t<t*)
Upper tailed test P-value = P(t>t*)
Two-tailed test P-value = 2P(t > |t*| )=2P(t <- |t*| )

In each case, you can reject H0 if P-value   and fail to reject H0 (accept H0) if P-
value > 

(b) Rejection region for level  test:

Lower tailed test t  -t;n-1
Upper tailed test t  t;n-1
Two- tailed test t  -t/2;n-1 or t  t/2;n-1

100(1-)% confidence Intervals with the same assumptions,

_
sD
d  t / 2;n 1
n

Example: The manager of a fleet of automobiles is testing two brands of radial tires. He assigns
one tire of each brand at random to the two rear wheels of eight cars and runs the cars until the
tires wear out. The descriptive statistics for the data are shown below (in kilometers). Find the
99% confidence interval on the difference in mean life. Which brand would you prefer based on
this calculation? Is there an alternative method to answer this question instead of computing the
confidence interval?

Variable N Mean Median StDev SE Mean Minimum Maximum Q1 Q3

Brand1 8 38479 37067 5590 1976 32100 48360 34185 43525
Brand2 8 37611 36655 5244 1854 31950 47800 33491 41214
Difference 8 868 475 1290 456.1 -805 3020 N/A N/A

Brand1 36925 45300 36240 32100 37210 48360 38200 33500

Brand2 34318 42280 35500 31950 38015 47800 37810 33215
Difference 2607 3020 740 150 -805 560 390 285

Exercise 9.44:
H 0 : 1   2  500 or H 0 : 1   2  500
H a : 1   2  500 or H a : 1   2  500
(a) the difference distribution should be normal.
(b) Computed differences, d=X-Y: 815 1084 681 550 535 786 1162 517 910
_
which gives the sample mean, d =782.2222 and the standard deviation, s D =236.736.
_
d  0 782.2222  500
Test statistics: t   =3.5764
sD / n 236.736 / 9
Since this is an upper tailed test, reject H 0 if t  t;n-1= t0.05;8 =1.86 or if the P-
value=P(t>3.5764) is less than =0.05. Notice that test statistics, 3.5764 is large than 1.86.
Also the 0.005 <P-value < 0.001 by looking at the table. Reject H 0 and there is sufficient
evidence to conclude that the true average yield in winter wheat is more than 500 higher than for
spring wheat.