CHAPTER 10

10.1 df = n1 + n2 2 = 12 + 15 2 = 25
10.3 Assume that you are sampling from two independent normal distributions having eual varian!es.
10.5 df =
1 2
2 n n + = 5 + 4 2 = 7
10." #a$ H
0
:
1 2
The mean estimated amount of calories in the cheeseburger is not lower
for the people who thought about the cheesecake first than for the people who thought
about the organic fruit salad first.
H
1
:
1 2
< The mean estimated amount of calories in the cheeseburger is lower for
the people who thought about the cheesecake first than for the people who thought about
the organic fruit salad first.
(b) Type I error is the error made in concluding that the mean estimated amount of calories
in the cheeseburger is lower for the people who thought about the cheesecake first than
for the people who thought about the organic fruit salad first when the mean estimated
amount of calories in the cheeseburger is in fact not lower for the people who thought
about the cheesecake first than for the people who thought about the organic fruit salad
first.
(c) Type II error is the error made in concluding that the mean estimated amount of calories
in the cheeseburger is not lower for the people who thought about the cheesecake first
than for the people who thought about the organic fruit salad first when the mean
estimated amount of calories in the cheeseburger is in fact lower for the people who
thought about the cheesecake first than for the people who thought about the organic fruit
salad first.
(d)
.
( ) ( )
2
2
2
1
2
1
2 1 2 1
n
S
n
S
X X
t
STAT
+

=

= -6.1532
Decision: Since t
STAT
= -6.1532 is smaller than the critical bound of -2.4314, reject H
0
.
There is evidence that the mean estimated amount of calories in the cheeseburger is
lower for the people who thought about the cheesecake first than for the people who
thought about the organic fruit salad first.
10.% #a$
0 1 2
& H = Mean times to clear problems at Office I and Office II are the same.
1 1 2
& H Mean times to clear problems at Office I and Office II are different.
Since the p-value of 0.725 is greater than the 5% level of significance, do not reject the
null hypothesis. There is not enough evidence to conclude that the mean time to clear
problems in the two offices is different.
(b) p-value = 0.725. The probability of obtaining a sample that will yield a t test statistic
more extreme than 0.3544 is 0.725 if, in fact, the mean waiting times between Office 1
and Office 2 are the same.
(c) We need to assume that the two populations are normally distributed.
(d) ( ) ( )
2
1 2
1 2
1 1 1 1
2.21' 2.0115 2.02'' 3.2(51
20 20
p
X X t S
n n


+ + = + +



1 2
0.%5'3 1.35%3
)in!e the *onfiden!e +nterval !ontains 0, we !annot !laim that there-s a differen!e between
the two means.
10.11 #a$ ./)tat output&
H
0
:
1
=
2
H
1
:
1

2
where .opulations& 1 = sub!ompa!t, 2 = !ompa!t
0e!ision rule& +f 1tSTAT 1 2 2.11%% or p3value 4 0.05, re5e!t H0.
( ) ( )

+

=
2 1
2
2 1 2 1
1 1
n n
S
X X
t
p
STAT

= 30.(161, p3value = 0.5'52
0e!ision& )in!e p3value 2 0.05, do not re5e!t H0. 7here is not enough eviden!e of a
differen!e in the mean battery life between the two types of digital !ameras.
#b$ p3value = 0.5'52. 7he probability of obtaining a sample that yields a t test statisti! farther
away from 0 in either dire!tion is 0.5'52 if there is no difference in the mean battery
life between the two types of digital cameras.
#!$
( ) ( )
2
1 2 8 2
1 2
1 1 1 1
20".2"2" 220 2.11%% 1613.(3('
11 "
p
X X t S
n n

+ = +



35(.3""1
2 1
30.%225
9ou are %5: !onfident that the difference between the population mean battery life
of the two types of digital cameras is somewhere between -56.3771 and
30.%225.
10.13
0 1 2
& H = Mean waiting times of Bank 1 and Bank 2 are the same.
1 1 2
& H Mean waiting times of Bank 1 and Bank 2 are different.
Since the p-value of 0.00031 is less than the 5% level of significance, reject the null
hypothesis. There is enough evidence to conclude that the mean waiting times are
different in the two banks.
Both t tests yield the same conclusion.
10.15 H
0
:
1
=
2
H
1
:
1

2

Decision: Since t
STAT
= 4.104 is greater than the upper critical bounds 2.0281, reject
H
0
. There is evidence of a difference in the mean surface hardness between untreated and
treated steel plates.
The value of pooled-variance t test statistic and the separate-variance t test statistic are
almost identical while the critical bound of the pooled-variance t test is slightly smaller
than that of the separate-variance t test because the degrees of freedom of the pooled-
variance t test is two more than that of the separate-variance t test.
10.1" #a$ H0&
1 2
= where .opulations& 1 = 7e!hnology, 2 = ;inan!ial +nstitutions
H1&
1 2

0e!ision rule& +f 1tSTAT 1 2 2.01(" or p3value 4 0.05, re5e!t H0.
7est statisti!&

+
=
2 1
2
2 1 2 1
1 1
$ # $ #
n n
S
X X
t
p
STAT

= 2.(50% p3value = 0.0112
0e!ision& )in!e p3value 4 0.05, re5e!t H0. 7here is enough eviden!e of a differen!e between
the te!hnology se!tor and the finan!ial institutions se!tor with respe!t to mean brand value.
#b$H0&
1 2
= where .opulations& 1 = 7e!hnology, 2 = ;inan!ial +nstitutions
H1&
1 2

0e!ision rule& +f 1tSTAT1 2 2.06(0 or p-value 4 0.05, re5e!t H0.
)in!e p3value 4 0.05, re5e!t H0. 7here is enough eviden!e of a differen!e between the
te!hnology se!tor and the finan!ial institutions se!tor with respe!t to mean brand value.
#!$ 7he !on!lusions in #a$ and #b$ are the same. )in!e the sample varian!e of the te!hnology
se!tor is more than 1% times as big as that of the finan!ial institutions se!tor, the test in #b$
is the appropriate test to perform assuming that both samples are drawn from normally
distributed populations.
10.1% d.f. = n 1 = 15 1 = 1', where n = number of pairs of data
10.21 #a$
0 1
& 0 vs. & 0
D D
H H =
7est statisti!&
n
S
D
t
D
D
STAT

=
= 15."3%( p3value = 0.0000
0e!ision& )in!e p3value is virtually <ero, re5e!t
0
H . 7here is eviden!e of a
differen!e in the mean !ellphone servi!e rating between =eri<on and A7>7.
#b$ 9ou must assume that the distribution of the differen!es between the mean measurements is
appro?imately normal.
#!$
7he normal probability plot suggests that the distribution is normal.
#d$
3.633'
12.6(3( 2.0"%(
22
D
S
D t
n
=
11.1(
D
1'.5(
9ou are %5: !onfident that differen!e in the mean !ellphone servi!e rating between =eri<on
and A7>7 is somewhere between 11.1( and 1'.5(.
10.23 (a) Define the difference to be the global rating minus the U.S. rating.
0 1
& 0 vs. & 0
D D
H H =
Test statistic:
n
S
D
t
D
D
STAT

=
= 30.0(6% p3value = 0.%'(2
Decision: Since p-value > 0.05, do not reject
0
H . There is not enough evidence of a
difference in the mean rating between global and U.S. employees.
(b) The differences are assumed to be normally distributed.
(c) The normal probability plot do not indicate severe departure from normality.
10.25
;rom the des!riptive statisti!s provided in the @i!rosoft A?!el output there does not seem to be
any violation of the assumption of normality. 7he mean and median are similar and the sBewness
value is near 0. Cithout observing other graphi!al devi!es su!h as a stem3and3leaf display,
bo?plot, or normal probability plot, the fa!t that the sample si<e #n = 35$ is not very small enables
us to assume that the paired t test is appropriate here.
7he @i!rosoft A?!el output for the paired t test indi!ates that a signifi!ant improvement in mean
performan!e ratings has o!!urred. 7he !al!ulated t statisti! of 2.(%% falls far below the one3tailed
!riti!al value of 1.(%0% using a 0.05 level of signifi!an!e. 7he p3value is 0.005'.
10.2" #a$
1 2
1 2
1 2
50 30
0.50, 0.30,
100 100
X X
p p
n n
= = = = = =
and p =
X
1
+ X
2
n
1
+ n
2
=
50 + 30
100 +100
= 0.'0
H0&
1
=
2
H1&
1

2

0e!ision rule& +f ZSTAT 4 1.%( or ZSTAT 2 1.%(, re5e!t H0.

7est statisti!&
( ) ( )
( )
( )
( )

+

=
100
1
100
1
0.'0 3 1 0.'0
0 0.30 3 0.50
1 1
1
2 1
2 1 2 1
n n
p p
p p
Z
STAT

= 2.6%
0e!ision& )in!e ZSTAT = 2.6% is above the !riti!al bound of 1.%(, re5e!t H0. 7here is
suffi!ient eviden!e to !on!lude that the population proportions differ for group 1 and group
2.
#b$
( )
( ) ( ) ( ) ( )
1 1 2 2
1 2
1 2
1 1 .5 .5 .3 ."
0.2 1.%( +
100 100
p p p p
p p Z
n n

+ =


1 2
0.0("1 0.332%
10.2% #a$
H0&
1
=
2
H1&
1

2
where .opulations& 1 = males, 2 = females
0e!ision rule& +f ZSTAT 4 2.56 or ZSTAT 2 2.56, re5e!t H0.
STAT
Z = 32.261"
0e!ision& )in!e ZSTAT = 32.261" is between the two !riti!al bounds, do not re5e!t H0. 7here
is insuffi!ient eviden!e to !on!lude that a signifi!ant differen!e e?ists in the proportion of
males and females who en5oy shopping !lothing for themselves.
#b$ p-value = 0.0225. 7he probability of obtaining a differen!e in two sample proportions of
0.0(%2 or more in either dire!tion when the null hypothesis is true is 0.0225.
#!$ ./)tat output&
( )
( ) ( ) ( ) ( )
1 1 2 2
1 2
1 2
1 1 0.'3%1 1 0.'3%1 0.5063 13 0.5063
0.0(%2 2.5"56 +
5'2 5'3
p p p p
p p Z
n n

+ =


30.1'"1
1 2
0.006"
9ou are %%: !onfident that the differen!e in the proportions of males and females who
en5oy shopping !lothing for themselves is between 30.1'"1 and 0.006".
#d$ #a$
H0&
1
=
2
H1&
1

2
where .opulations& 1 = males, 2 = females
0e!ision rule& +f ZSTAT 4 2.56 or ZSTAT 2 2.56, re5e!t H0.
STAT
Z = 33.5060
0e!ision& )in!e ZSTAT = 33.5060 is less than the lower !riti!al bound, re5e!t H0.
7here is suffi!ient eviden!e to !on!lude that a signifi!ant differen!e e?ists in the
proportion of males and females who en5oy shopping !lothing for themselves.
#b$ p-value = 0.0005. 7he probability of obtaining a differen!e in two sample
proportions of 0.10(1 or more in either dire!tion when the null hypothesis is true
is 0.0005.
#!$
( )
( ) ( ) ( ) ( )
1 1 2 2
1 2
1 2
1 1 0.'022 1 0.'022 0.5063 13 0.5063
0.10(1 2.5"56 +
5'2 5'3
p p p p
p p Z
n n

+ =


30.1635
1 2
30.026(
9ou are %%: !onfident that the differen!e in the proportions of males and females
who en5oy shopping !lothing for themselves is between 30.1635 and 30.026(.
10.31 #a$
1
351
3('2
p = = 0.0%('
#b$
2
'65
355(
p = = 0.13('
#!$ H0&
1

2
H1&
1
<
2

where .opulations& 1 = original !all to a!tion button, 2 = new !all to a!tion button
0e!ision rule& +f ZSTAT 4 1.(''% or p3value 4 0.05, re5e!t H0.
7est statisti!&
( ) ( )
( )
1 2 1 2
1 2
1 1
1
STAT
p p
Z
p p
n n

=

+


= -5.2974 p-value = 0.0000
0e!ision& )in!e p3value 4 0.05, re5e!t H0. 7here is eviden!e that the new !all to a!tion
button is more effe!tive than the original.
10.33 (a) PHStat output:
H
0
:
1

=
2
H
1
:
1

2

Decision rule: If |Z
STAT
| > 1.96, reject H
0
.
Test statistic:
1 2
1 2
2"3 252
((5 500
X X
p
n n
+ +
= =
+ +
= 0.4506
( ) ( )
( )
( )
( )
1 2 1 2
1 2
0.'105 0.50' 0
1 1
1 1
0.'50( 1 0.'50(
1
((5 500
STAT
p p
Z
p p
n n

= =

+
+




= 3.1738
Decision: Since |Z
STAT
| = 3.1738 is larger than 1.96, reject H
0
. There is sufficient evidence
of a difference between consumer magazines and newspapers in the proportion of online-
only content that is copy-edited as rigorously as print content.
(b) p-value = 0.0015. The probability of obtaining a difference in proportions that gives rise
to a test statistic that is 3.1738 or more away from 0 in either direction is 0.0015 if there
is not a difference between consumer magazines and newspapers in the proportion of
online-only content that is copy-edited as rigorously as print content.
(c) H
0
:
1

=
2
H
1
:
1

2

Decision rule: If |Z
STAT
| > 1.96, reject H
0
.
Test statistic:
1 2
1 2
3"% 2%(
((5 500
X X
p
n n
+ +
= =
+ +
= 0.5794
( ) ( )
( )
( )
( )
1 2 1 2
1 2
0.5(%% 0.5%2 0
1 1
1 1
0.5"%' 1 0.5"%'
1
((5 500
STAT
p p
Z
p p
n n

= =

+
+




= -0.7555
Decision: Since |Z
STAT
| = 0.7555 is smaller than 1.96, do not reject H
0
. There is not
sufficient evidence of a difference between consumer magazines and newspapers in the
proportion of online-only content that is fact-checked as rigorously as print content.
(d) p-value = 0.45. The probability of obtaining a difference in proportions that gives rise to
a test statistic that is 0.7555 or more away from 0 in either direction is 0.45 if there is not
a difference between consumer magazines and newspapers in the proportion of online-
only content that is fact-checked as rigorously as print content.
10.35 (a) H
0
:
1
=
2
where Populations: 1 = under age 50, 2 = age above 50
H
1
:
1

2

Decision rule: If |Z
STAT
| > 1.96, reject H
0
.
Test statistic:
( ) ( )
( )
1 2 1 2
1 2
1 1
1
STAT
p p
Z
p p
n n

= =

+


14.8797
Decision: Since |Z
STAT
| = 14.8797 is greater than 1.96, reject H
0
. There is sufficient
evidence of a significant difference in the proportion of users under age 50 and users 50
years and older that accessed the news on their cellphones.
(b) p-value is virtually 0. The probability of obtaining a difference in proportions that gives
rise to a test statistic that deviates from 0 in either direction by 14.8797 or more in either
direction is virtually 0 if there is no difference in the proportion of users under age 50
and users 50 years and older that accessed the news on their cellphones.
(c)
( )
( ) ( )
1 1 2 2
1 2
1 2
1 1 p p p p
p p Z
n n

+
( )
( ) ( ) 0.'" 130.'" 0.150' 1 0.150'
0.'" 0.150' 1.%(
1000 6%1

= +
0.2808
2 1
0.3584
10.3" #a$

=0.05,
1
n =16,
2
n =21,
0.05
F = 2.20
#b$

=0.01,
1
n =16,
2
n =21,
0.01
F = 3.0%
10.3%
" . 133
% . 1(1
2
2
2
1
= =
S
S
F
STAT
= 1.210%
10.'1

=0.05,
1
n =25,
2
n =25,
0.058 2
F = = 2.2"
10.'3 +n testing the euality of two population varian!es, the F3test statisti! is very sensitive to the
assumption of normality for ea!h population. +f the populations are very right3sBewed, the F3test
should not be used.
10.'5 #a$ H0&
2 2
1 2
= where .opulations& 1 = Dine 2, 2 = Dine 1
H1&
2 2
1 2

0e!ision rule& +f FSTAT 2 2.52(5, re5e!t H0. 7est statisti!&
2
1
2
2
STAT
S
F
S
= = 1.212'
0e!ision& )in!e FSTAT = 1.212' is less than the !riti!al bound of

F
= 2.52(5, do not
re5e!t H0. 7here is not enough eviden!e of a differen!e in the variability of the time to !lear
problems between the two !entral offi!es.
#b$ p-value = 0.("6%
7he probability of obtaining a sample that yields a test statisti! more e?treme than 1.212'
is 0.("6% if the null hypothesis that there is no differen!e in the two population varian!es is
true.
#!$ 7he test assumes that the two populations are both normally distributed.
#d$ Eased on #a$ and #b$, a pooled3varian!e t test should be used.
10.'" #a$ H0&
1
2
=
2
2
7he population varian!es are the same.
H1&
1
2

2
2
7he population varian!es are different.
0e!ision rule& +f FSTAT 2 2.%"6(, re5e!t H0.
7est statisti!&
2
2
2
2
2
1
(360 . 1
0622 . 2
= =
S
S
F
STAT
= 1.(15%
0e!ision& )in!e FSTAT = 1.(15% is below the upper !riti!al bound of
2 8
F
= 2.%"6(, do not
re5e!t H0. 7here is not enough eviden!e to !on!lude that the two population varian!es are
different.
#b$ p3value = 0."15. 7he probability of obtaining a sample that yields a test statisti! more
e?treme than 1.(15% is 0."15 if the null hypothesis that there is no differen!e in the two
population varian!es is true.
#d$ Eased on the results of #a$, it is appropriate to use the pooled3varian!e t3test to !ompare
the means of the two bran!hes.
10.'% #a$ H0&
1
2
=
2
2
where .opulations& 1 = males,
2 = females
H1&
1
2

2
2
0e!ision rule& +f FSTAT 2 2.1010, re5e!t H0.
7est statisti!&
2
1
2
2
STAT
S
F
S
= = 12.("(5
0e!ision& )in!e FSTAT = 12.("(5 is greater than the upper !riti!al bound of
2 8
F = 2.1010,
re5e!t H0. 7here is eviden!e of a differen!e in the varian!es of time spent on ;a!ebooB per
day between males and females.
#b$ Assuming the underlying normality in the two populations is met, based on the results
obtained in part #a$, it is more appropriate to use the separate3varian!e t3test.
10.51 Among the !riteria to be used in sele!ting a parti!ular hypothesis test are the type of data, whether
the samples are independent or paired, whether the test involves !entral tenden!y or variation,
whether the assumption of normality is valid, and whether the varian!es in the two populations are
eual.
10.53 7he F test !an be used to e?amine differen!es in two varian!es when ea!h of the two populations is
assumed to be normally distributed.
10.55 Fepeated measurements represent two measurements on the same items or individuals, while paired
measurements involve mat!hing items a!!ording to a !hara!teristi! of interest.
10.5" Chen you have obtained data from either repeated measurements or paired data.
10.5% #a$ H
0
:
1
2
=
2
2
The population variances are the same.
H
1
:
1
2

2
2
The population variances are different.
Decision rule: If F
STAT
> 2.1169, reject H
0
.
Test statistic:
( )
( )
2
2
1
2 2
2
5."1''
5.'0('
STAT
S
F
S
= = = 1.1172
Decision: Since F
STAT
= 1.1172 is smaller than the upper critical bound of 2.1169, do not
reject H
0
. There is not enough evidence of any difference in the variance of the study
time for male students and female students.
#b$ )in!e there is not enough eviden!e of any differen!e in the varian!e of the study time for
male students and female students, a pooled3varian!e t test should be used.
#!$ H
0
:
2 1
= H
1
:
1

2
Decision rule: d.f. = 56. If t
STAT
< 2.0032 or t
STAT
> 2.0032, reject H
0
.
Decision: Since t
STAT
= 3.6762 is larger than the upper critical bound of 2.0032, reject H
0
.
(d) There is enough evidence of a difference in the mean study time for male and female
students.
10.61
H
0
:
2
2
2
1
= H
1
:
2
2
2
1
Let Population 1 = suburban, 2 = city
Since the p-value = 0.3302 > 0.05, do not reject
0
H . At 5% level of significance, there is
insufficient evidence to conclude that the two variances are not the same. Hence, a pooled
variance t test is more appropriate.
H
0
:
2 1
= H
1
:
1

2
Since the p-value = 0.6441 is greater than the 5% level of significance, do not reject
0
H .
There is not enough evidence to conclude that the mean food rating between suburban and city
restaurants is different.
H
0
:
2
2
2
1
= H
1
:
2
2
2
1
Let Population 1 = suburban , 2 = city
Since the p-value = 0.5325 is greater than the 5% level of significance, do not reject
0
H . There is
not enough evidence to conclude that the variances are different. Hence, a pooled-variance t test
is appropriate.
H
0
:
2 1
= H
1
:
1

2
Since the p-value = 0.9241 is greater than the 5% level of significance, do not reject
0
H . There is
not enough evidence to conclude that the mean dcor rating between city restaurants and
suburban restaurants is different.
H
0
:
2
2
2
1
= H
1
:
2
2
2
1
Let Population 1 = suburban, 2 = city
Since the p-value = 0.6685 is greater than the 5% level of significance, do not reject
0
H . There is
not enough evidence to conclude that the variances are different. Hence, a pooled-variance t test
is appropriate.
H
0
:
2 1
= H
1
:
1

2
Since the p-value = .4155 is larger than the 5% level of significance, do not reject
0
H .
There is not enough evidence to conclude that the mean service rating between city restaurants
and suburban restaurants is different.
H
0
:
2
2
2
1
= H
1
:
2
2
2
1
Let Population 1 = city, 2 = suburban
Since the p-value = 0.0000 is smaller than the 5% level of significance, reject
0
H . There is
enough evidence to conclude that the variances are different. Hence, a separate-variance t test is
appropriate.
H
0
:
2 1
= H
1
:
1

2
Since the p-value = 0.0005 is smaller than the 5% level of significance, reject
0
H .
There is enough evidence to conclude that the mean price between city restaurants and suburban
restaurants is different.
10.63 Population 1 = men, 2 = women
(a) H
0
:
1
2
=
2
2
The population variances are the same.
H
1
:
1
2

2
2
The population variances are different.
Since the p-value = 0.0019 is lower than the 5% level of significance, reject
0
H . There is
enough evidence of a difference in the variances of the number of online friends between
men and women. Hence, a separate-variance t test is appropriate.
10. 63 (b) H
0
:
2 1
= H
1
:
1

2
(b) Since the p-value is virtually zero, reject
0
H . There is enough evidence of a difference
in the mean number of online friends between men and women.
(c)
( ) ( )
2
1 2 8 2
1 2
1 1 1 1
16031'0 1.%(06 15(5'.5"
1511 1500
p
X X t S
n n

+ = +



31.0563
2 1
'6.%'1"
10.65 Population 1 = Wing A, 2 = Wing B
H
0
:
1
2
=
2
2
The population variances are the same.
H
1
:
1
2

2
2
The population variances are different.
Decision rule: If F
STAT
> 2.5265, reject H
0
.
Test statistic:
( )
( )
2
2
2
2
2
1
1.3"00
1.'1"2
= =
S
S
F
STAT
= 1.0701
Decision: Since F
STAT
= 1.0701 is lower than the critical bound of
2 8
F = 2.5265, do not reject
H
0
. There is not enough evidence to conclude that there is a difference between the variances in
Wing A and Wing B. Hence, a pooled-variance t test is more appropriate for determining
whether there is a difference in the mean delivery time in the two wings of the hotel.
H
0
:
2 1
= H
1
:
1

2
Decision rule: d.f. = 38. If t
STAT
< 2.0244or t
STAT
> 2.0244, reject H
0
.
Test statistic:
( ) ( )
( ) ( ) 1 1
1 1
2 1
2
2 2
2
1 1
2
+
+
=
n n
S n S n
S
p
=
( )( ) ( ) ( )
( ) ( ) 1 20 1 20
1.'1"2 1 20 1.3"00 1 20
2 2
+
+
= 1.9427
( ) ( )

+

=
2 1
2
2 1 2 1
1 1
n n
S
X X
t
p
STAT

=
( )

20
1
20
1
1.%'2"
0 6.12 3 10.'0
= 5.1615
Decision: Since t
STAT
= 5.1615 is greater than the upper critical bound of 2.0244, reject H
0
. There
is enough evidence of a difference in the mean delivery time in the two wings of the hotel.
10.67
0
&
B V
H = Mean weights of Boston and Vermont shingles are the same.
1
&
B V
H
Mean weights of Boston and Vermont shingles are different.
Since the p-value is essentially zero, reject
0
H . There is sufficient evidence to conclude that the
mean weights of Boston and Vermont shingles are different.
10.69 Since the sample size is small, you have to assume that the 3-year return, 5-year return, 10-year
return and expense ratio are all normally distributed to perform the following tests.
1-year return:
Populations: 1 = long-term, 2 = short-term
H
0
:
1
2
=
2
2
The population variances are the same.
H
1
:
1
2

2
2
The population variances are different.
PHstat output:
Decision: Since p-value < 0.05, reject H
0
. There is enough evidence to conclude that the
two population variances are different. Hence, the appropriate test for the difference in
two means is the separate-variance t test.
10.69 Populations: 1 = long-term, 2 = short-term
H
0
:
1 2
= H
1
:
1 2

Decision: Since the p-value = 0.0010 is less than 0.05, reject H
0
. There is sufficient
evidence to conclude that the mean 1-year return is different between the long-term and
short-term bond funds.
10.69 3-year return:
Populations: 1 = long-term, 2 = short-term
H
0
:
1
2
=
2
2
The population variances are the same.
H
1
:
1
2

2
2
The population variances are different.
PHstat output:
Decision: Since p-value > 0.05, do not reject H
0
. There is not enough evidence to
conclude that the two population variances are different. Hence, the appropriate test for
the difference in two means is the pooled-variance t test.
10.69 Populations: 1 = long-term, 2 = short-term
H
0
:
1 2
= H
1
:
1 2

PHStat output:
Decision: Since the p-value < 0.05, reject H
0
. There is sufficient evidence to conclude
that the mean 3-year return is different between the long-term and short-term funds.