8-1

Introductory Statistical
Methods
Preparatory Module
 8-2

Chapter 8

The Comparison of
Two Populations
 8-3

8 The Comparison of Two Populations

• Using Statistics
• Paired-Observation Comparisons
• A Test for the Difference between Two Population
Means Using Independent Random Samples
• A Large-Sample Test for the Difference between
Two Population Proportions
• The F Distribution and a Test for the Equality of
Two Population Variances
 8-4

8 LEARNING OBJECTIVES
After studying this chapter you should be able to:
• Explain the need to compare two population parameters
• Conduct a paired difference test for the difference in
population means
• Conduct an independent samples test for the difference in
population means
• Describe why a paired difference test is better than
independent samples test
• Conduct a test for difference in population proportions
• Test whether two population variances are equal
 8-5

8-1 Using Statistics

• Inferences about differences between

parameters of two populations
 Paired-Observations

 Observe the same group of persons or things

 At two different times: “before” and “after”
 Under two different sets of circumstances or “treatments”
 Independent Samples
 Observe different groups of persons or things
 At different times or under different sets of circumstances
 8-6

8-2 Paired-Observation Comparisons

• Population parameters may differ at two

different times or under two different sets of
circumstances or treatments because:
 The circumstances differ between times or treatments
 The people or things in the different groups are
themselves different
• By looking at paired-observations, we are
able to minimize the “between group” ,
extraneous variation.
 8-7

Paired-Observation Comparisons
of Means

Test statistic for the paired - observations t test :

Test statistic for the paired - observations t test :

D  D
t D   D0 0
t s
sD D
n
n
where D is the sample average difference between each
where D is the sample average difference between each
pair of observations, s is the sample standard deviation
pair of observations, s D Dis the sample standard deviation
of these differences, and the sample size, n, is the number
of these differences, and the sample size, n, is the number
of pairs of observations. The symbol  is the population
of pairs of observations. The symbol  D0D0is the population
mean difference under the null hypothesis. When the null
mean difference under the null hypothesis. When the null
hypothesis is true and the population mean difference is  ,
hypothesis is true and the population mean difference is  D0D,0
the statistic has a t distributi on with (n - 1) degrees of freedom.
the statistic has a t distributi on with (n - 1) degrees of freedom.
 8-8

Example 8-1

AArandom
randomsample
sampleofof1616viewers
viewersofofHome
HomeShopping
ShoppingNetwork
Networkwaswasselected
selectedfor
foran
anexperiment.
experiment. AllAllviewers
viewersinin
thesample
the samplehad
hadrecorded
recordedthe
theamount
amountofofmoney
moneythey
theyspent
spentshopping
shoppingduring
duringthe
theholiday
holidayseason
seasonofofthe
theprevious
previous
year. The
year. Thenext
nextyear,
year,these
thesepeople
peoplewere
weregiven
givenaccess
accesstotothe
thecable
cablenetwork
networkand
andwere
wereasked
askedtotokeep
keepaarecord
recordofof
theirtotal
their totalpurchases
purchasesduring
duringthe
theholiday
holidayseason.
season. Home
HomeShopping
ShoppingNetwork
Networkmanagers
managerswant
wanttototest
testthe
thenull
null
hypothesisthat
hypothesis thattheir
theirservice
servicedoes
doesnot
notincrease
increaseshopping
shoppingvolume,
volume,versus
versusthe
thealternative
alternativehypothesis
hypothesisthat thatititdoes.
does.

Shopper Previous Current Diff

Shopper Previous Current
1
1 334
334
405
405
Diff
71
71
HH[Link]DD0
0
2
2
3
150
150
520
125
125
540
-25
-25
20
HH[Link]DD>>00
3 520 540 20
4 95 100 5
4 95 100 5
5 212 200 -12
5
6
212
30
200
30
-12
0 dfdf==(n-1)
(n-1)==(16-1)
(16-1)==15
15
D 
6 30 30 0 D
7 1055 1200 145
7 1055 1200 145 0
8
8 300
300
265
265
-35
-35 Statistic: t 
TestStatistic:
Test
9 85 90 5 sD
9 85 90 5
10 129 206 77
10 129 206 77
11 40 18 -22 n
11 40 18 -22
12 440 489 49
12 440 489 49
13
13
610
610
590
590
-20
-20
CriticalValue:
Critical Value:t t0.05==1.753
1.753
0.05
14 208 310 102
14 208 310 102
15 880 995 115
15 880 995 115
16
16
25
25
75
75
50
50 Donot
Do notreject 1.753
rejectHH0ifif: :t t1.753
0
RejectHH0if:
Reject if:t t >>1.753
1.753
0
 8-9

Example 8-1: Solution

t t2.354
t t== =2.354
2.354
=2.354 >1.753,
1.753,
>>1.753,
>1.753, so
soso
so HHH H rejected
isrejected
0 is0is
rejectedand and
and we
wewe conclude
conclude
conclude
0 is0 rejected and we conclude that
that
that
that
D D  there
there isevidence
evidence that shopping volume by network
D D 3232
0 0 . 81 . 
81 0 0 there
there isisevidence
isevidence that
that
that shopping
shopping
shopping volume
volume
volume byby
by network
network
network
t t  s s    2.354
2.354 viewers
viewers
viewers
viewers has has
has
has increased,
increased,
increased,
increased, with
with with
witha ap-value
p-value
ap-value
ap-value between
between
between
between 0.01
0.01
0.01
0.01 an
anan
an
DD 55 55
. 75 . 75
0.025.
0.025.
0.025.
0.025. The
The The
The Template
Template
Template
Template output
output
output
output gives
gives
gives
gives amore
aamore more
amore exact
exact
exact
exact p-value
p-value
p-value
p-value
n n 1616
ofofofof 0.0163.
0.0163.
0.0163.
0.0163. See
See See
See the
the the
the next
next next
next slide
slide
slide
slide for
for for
for the
the the
the output.
output.
output.
output.
t Distribution:
t Distribution: df=15
df=15
0.4 0.4

0.3 0.3
f(t)
f(t)

0.2 0.2
Nonrejection
Nonrejection Rejection
Rejection
Region
Region Region
Region
0.1 0.1

0.0 0.0
-5 -5 0 1.753
01.753 5 5 t t
= t0.05= t0.05
2.131 2.131 2.602
2.602
= t0.025
= t0.025 =t =t
0.01 0.01

2.354=
2.354=
test test
statistic
statistic
 8-10

Example 8-2

ItIthas
hasrecently
recentlybeen
beenasserted
assertedthat
thatreturns
returnsononstocks
stocksmaymaychange
changeonceonceaastory
storyabout
aboutaacompany
companyappears
appearsininThe
TheWall
Wall
StreetJournal
Street Journalcolumn
column“Heard
“Heardononthe
theStreet.”
Street.” AnAninvestments
investmentsanalyst
analystcollects
collectsaarandom
randomsample
sampleofof5050stocks
stocksthat
that
wererecommended
were recommendedasaswinners
winnersbybythe
theeditor
editorofof “Heard
“Heardononthe
theStreet,”
Street,”and
andproceeds
proceedstotoconduct
conductaatwo-tailed
two-tailedtest
test
ofofwhether
whetherorornotnotthe
theannualized
annualizedreturn
returnon
onstocks
stocksrecommended
recommendedininthe thecolumn
columndiffers
differsbetween
betweenthethemonth
monthbefore
before
andthe
and themonth
monthafter
afterthe
therecommendation.
recommendation. For Foreach
eachstock
stockthe
theanalysts
analystscomputes
computesthethereturn
returnbefore
beforeand
andthe
thereturn
return
afterthe
after theevent,
event,and
andcomputes
computesthe
thedifference
differenceininthe
thetwo
tworeturn
returnfigures.
figures. He
Hethen
thencomputes
computesthe theaverage
averageand
andstandard
standard
deviationofofthe
deviation thedifferences.
differences.

H0: D 0
H1: D > 0 DD D
D 0 00.1.100 14.14
zz ss 0  0.05 14 .14
DD 0 . 05
n = 50 nn 5050

D = 0.1% p - value : p ( z  14.14)  0

p - value : p ( z  14.14)  0
sD = 0.05%
D  Thistest
This testresult
resultisishighly
highlysignifican
significant,t,
D
z 0 andHH maymaybeberejected
rejectedatatany
anyreasonable
reasonable
Test Statistic: s and 0
D 0
n level ofsignifican
level of significance. ce.
 8-11

Confidence Intervals for Paired

Observations

(1--))100%
AA(1 100%confidence
confidenceinterval
intervalfor
forthe
themean difference ::
meandifference
DD

D  t ssDD
D  t 2 n
 2
n
wherett2 2isisthe
where thevalue
valueof
of the
thettdistributi
distributiononwith
with(n(n--1)
1)degrees
degreesof
offreedom
freedom
thatcuts
that cutsoff
offan anarea of 22totoits
areaof itsright,
right, When
Whenthe
thesample
samplesize
sizeisislarge,
large,
wemay
we mayapproximat
approximateett with zz . .
with

 2 2  22
 8-12

Confidence Intervals for Paired

Observations – Example 8-2

95%confidence
95% confidenceinterval
intervalfor
forthe
thedata
datain Example8822::
inExample
D  z ssDD  0.11.960.05 0.05  0.1 (1.96)(.0071)
D  z n 0.11.96
 2  0.1 (1.96)(.0071)
2
n 50
50
00.1.100.014
.014[[00.086
.086,0,0.114
.114]]

Notethat
Note thatthis
thisconfidence
confidenceinterval
intervaldoes
doesnot
notinclude
includethe
thevalue
value0.0.
 8-13

8-3 A Test for the Difference between Two Population

Means Using Independent Random Samples

• When paired data cannot be obtained, use

independent random samples drawn at
different times or under different
circumstances.
 Large sample test if:
 Both n130 and n230 (Central Limit Theorem), or
 Both populations are normal and 1 and 2 are both known
 Small sample test if:
 Both populations are normal and 1 and 2 are unknown
 8-14

Comparisons of Two Population

Means: Testing Situations
• I: Difference between two population means is 0
 1= 2
 H0: 1 -2 = 0
 H1: 1 -2  0
• II: Difference between two population means is less than 0
 12
 H0: 1 -2  0
 H1: 1 -2  0
• III: Difference between two population means is less than D
 1  2+D
 H0: 1 -2  D
 H1: 1 -2  D
 8-15

Comparisons of Two Population

Means: Test Statistic
Large-sampletest
Large-sample teststatistic
statisticfor
forthe
thedifference
differencebetween
betweentwo
two
populationmeans:
population means:

( x  x )  (   )
z 1 2 1 2 0

 2
 2
1
 2

n
1
n 2

Theterm
The term((1--2))0isisthe
thedifference between1an
differencebetween an2under
underthethe
1 2 0 1 2
nullhypothesis.
null hypothesis. IsIsisisequal
equaltotozero
zeroin
insituations
situationsIIand
andII,
II,and
andititisis
equalto
equal tothe
theprespecified
prespecifiedvaluevalueDDin insituation
situationIII.
III. The
Theterm
termin inthe
the
denominatorisisthe
denominator thestandard
standarddeviation
deviationof ofthe
thedifference
differencebetween
between
thetwo
the twosample
samplemeans
means(it (itrelies
relieson
onthe
theassumption
assumptionthatthatthe
thetwo
two
samplesare
samples areindependent).
independent).
 8-16

Two-Tailed Test for Equality of Two

Population Means: Example 8-3
Is there evidence to conclude that the average monthly charge in the entire population of American Express Gold
Card members is different from the average monthly charge in the entire population of Preferred Visa
cardholders?

Population 1 : Preferred Visa

H :  0
0 1 2
H :  0
1 1 2
n = 1200
1

x = 452
1
( x  x )  (   )
1 2 1 2 0  ( 452  523)  0
z 
 = 212
1

2

1  2
2
212
2

185
2

n n 1200 800
Population 2 : Gold Card 1 2
 71  71
   7.926
80.2346 8.96
n = 800
2

p - value : p(z < -7.926)  0

x = 523
2

 = 185
2 H is rejected at any common level of significan ce
0
 8-17

Example 8-3: Carrying Out the Test

Standard Normal Distribution Sincethe

Since thevalue
valueof ofthe
thetest
test
0.4
statisticisisfar
statistic farbelow
belowthe thelower
lower
0.3
criticalpoint,
critical point,the
thenull
null
hypothesismay
hypothesis maybe berejected,
rejected,
f(z)

0.2

0.1
andwe
and wemaymayconclude
concludethat that
0.0
0
z thereisisaastatistically
there statistically
-z0.01=-2.576 z0.01=2.576
Rejection Nonrejection Rejection significantdifference
significant differencebetween
between
Region Region Region
Test Statistic=-7.926 theaverage
the averagemonthly
monthlycharges
charges
ofGold
of GoldCardCardandandPreferred
Preferred
Visacardholders.
Visa cardholders.
 8-18

Two-Tailed Test for Difference Between

Two Population Means: Example 8-4
Is there evidence to substantiate Duracell’s claim that their batteries last, on average, at least 45 minutes longer
than Energizer batteries of the same size?

Population 1 : Duracell H :     45
0 1 2
H :     45
1 1 2
n = 100
1

( x  x )  (   )
x = 308
1 z 1 2 1 2 0  (308  254)  45
2 2 2 2
  84 67
 = 84
1
1  2 
n n 100 100
1 2
Population 2 : Energizer
9 9
   0.838
115.45 10.75

n = 100
2
p - value : p(z > 0.838) = 0.201
x = 254
2
H may not be rejected at any common
0
 = 67
2
level of significance
 8-19

Confidence Intervals for the Difference

between Two Population Means
AAlarge-sample
large-sample(1-)100%
(1-)100%confidence
confidenceinterval
intervalfor
forthe
thedifference
difference
betweentwo
between twopopulation means,1--2,,using
populationmeans, usingindependent
independentrandom
random
1 2
samples:
samples:
2 2
 
(x  x )  z 1  2
1 2  n n
1 2
2

AA95%
95%confidence
confidenceinterval
intervalusing
usingthe
thedata
datain
inexample
example8-3:
8-3:
2 2
  212 2 1852
(x  x )  z 1  2  (523  452)  1.96   [53.44,88.56]
1 2  n n 1200 800
1 2
2
 8-20

A Test for the Difference between Two Population

Means: Assuming Equal Population Variances

Ifwe
 If wemight
mightassume
assumethat
thatthe
thepopulation variances212and
populationvariances and222 are
areequal
equal
1 2
(eventhough
(even thoughunknown),
unknown),thenthenthe
thetwo
twosample
samplevariances,
variances,ss212and
andss22,2,
1 2
providetwo
provide twoseparate
separateestimators
estimatorsofofthe
thecommon
commonpopulation
populationvariance.
variance.
Combiningthe
Combining thetwo
twoseparate
separateestimates
estimatesinto
intoaapooled
pooledestimate
estimateshould
should
giveus
give usaabetter
betterestimate
estimatethan
thaneither
eithersample
samplevariance
varianceby
byitself.
itself.
Deviation from the Deviation from the
mean. One for each mean. One for each
sample data point. sample data point.
}

}
** * * * * **
** * * * * * ** * * ** * * ** * *
Sample 1 Sample 2
x1 x2
From sample 1 we get the estimate s12 with From sample 2 we get the estimate s22 with
(n1-1) degrees of freedom. (n2-1) degrees of freedom.

From both samples together we get a pooled estimate, sp2 , with (n1-1) + (n2-1) = (n1+ n2 -2)
total degrees of freedom.
 8-21

Pooled Estimate of the Population

Variance
AApooled
pooledestimate
estimateofofthe
thecommon
commonpopulation
populationvariance,
variance,based
basedon onaasample
sample
variancess212from
variance fromaasample
sampleofofsize
sizenn1and
andaasample
samplevariance
variancess222from
fromaasample
sample
1 1 2
ofsize
of sizenn2isisgiven
givenby:
by:
2
(n1  1) s1  (n2  1) s2
2 2

s 2p 
n1  n2  2

Thedegrees
The degreesof
offreedom
freedomassociated
associatedwith
withthis
thisestimator
estimatoris:
is:
df==(n
df (n+1+ n2-2)
1 n2-2)

Thepooled
The pooledestimate
estimateof ofthe
thevariance
varianceisisaaweighted
weightedaverage
averageof ofthe
thetwo
two
individualsample
individual samplevariances,
variances,with
withweights
weightsproportional
proportionaltotothe
thesizes
sizesof
ofthe
thetwo
two
samples. That
samples. Thatis,
is,larger
largerweight
weightisisgiven
giventotothe
thevariance
variancefrom
fromthethelarger
larger
sample.
sample.
 8-22

Using the Pooled Estimate of the

Population Variance
2 11 11 
sspp  
Theestimate
estimateof
ofthe
thestandard
standarddeviation
deviationof (x1xx 2))isisgiven
of(x givenby:
by: 2
nn11 nn22
The 1 2

Teststatistic
Test statisticfor
forthe
thedifference
differencebetween
betweentwotwopopulation
populationmeans,
means, assuming
assumingequal
equal
populationvariances:
population variances:
(x1xx 2))(( 1 2)) 0
(x1 2 1 2 0
t t==
22 11 11
sspp n  n 
n11 n22
where(( 1 2)) 0 isisthe
where thedifference
differencebetween
betweenthe
thetwo
twopopulation
populationmeans
meansunder
underthe
thenull
null
1 2 0
hypothesis(zero
hypothesis (zeroororsome
someother
othernumber
numberD).D).

Thenumber
The numberofofdegrees
degreesofoffreedom
freedomofofthe
thetest statisticisisdfdf ==((nn1nn2 22))(the
teststatistic (the
2 1 2
numberofofdegrees
degreesofoffreedom
freedomassociated
associatedwith 2
withss p, , the
thepooled
pooledestimate
estimateofofthe
the
number p
populationvariance.
population variance.
 8-23

Example 8-5
Dothe
Do thedata
dataprovide
providesufficient
sufficientevidence
evidencetotoconclude
concludethat
thataverage
averagepercentage
percentageincrease
increaseininthe
theCPI
CPIdiffers
differswhen
whenoil
oil
sellsatatthese
sells thesetwo
twodifferent
differentprices?
prices?

Population 1: Oil price = $27.50 H 0 : 1   2  0

n1 = 14 H1:  1   2  0
( x1  x 2 )  (  1   2 ) 0
x1 = 0.317% t 
s1 = 0.12%  ( n1  1) s12  ( n2  1) s22   1 1 
   
Population 2: Oil price = $20.00  n1  n2  2   n1 n2 
n2 = 9 0.107 0.107
   2.154
x 2 = 0.21% 0.00247 0.0497

s 2 = 0.11%
Critical point: t = 2.080
0.025
H 0 may be rejected at the 5% level of significance
df = (n  n  2 )  (14  9  2 )  21
1 2
 8-24

Example 8-6

The manufacturers of compact disk players want

to test whether a small price reduction is enough H :  0
0 2 1
to increase sales of their product. Is there H :  0
evidence that the small price reduction is enough 1 2 1
to increase sales of compact disk players? ( x  x )  (   )
t 2 1 2 1 0
2 2
 ( n  1) s  ( n  1) s  1 1 
Population 1: Before Reduction  1 1 2 2   
 n n 2  n n 
n 1 = 15  1 2  1 2 
x 1 = $6598 ( 6870  6598)  0

s1 = $844  (14)844 2  (11)669 2  1 1 
   
 15  12  2  15 12 
Population 2: After Reduction  
n 2 = 12 272 272
   0.91
89375.25 298.96
x 2 = $6870
s 2 = $669
Critical point : t = 1.316
0.10
df = (n  n  2 )  (15  12  2 )  25 H may not be rejected even at the 10% level of significance
0
1 2
 8-25

Example 8-6: Continued

t Distribution: df = 25 Sincethe
Since thetest
teststatistic
statisticisisless
less
thantt0.10,the
thenull
nullhypothesis
hypothesis
0.4

0.3
than 0.10,
cannotbe
cannot berejected
rejectedatatany
any
f(t)

0.2
reasonablelevel
reasonable levelof
of
0.1
significance. We
significance. Weconclude
conclude
thatthe
theprice
pricereduction
reductiondoesdoes
0.0
-5 -4 -3 -2 -1 0 1 2 3
t0.10=1.316
4 5 t
that
Nonrejection
Region
Rejection
Region
notsignificantly
not significantlyaffectaffectsales.
sales.
Test Statistic=0.91
 8-26

Confidence Intervals Using the Pooled

Variance
AA(1-)
(1-)100%
100%confidence
confidenceinterval
intervalfor
forthe
thedifference
differencebetween
betweentwo
two
population means,1--2,,using
populationmeans, usingindependent
independentrandom
randomsamples
samplesand
and
1 2
assumingequal
assuming equalpopulation
populationvariances:
variances:

2 1 
1
( x1  x2 )  t sp   
  1 2
n n
2

AA95%
95%confidence
confidenceinterval
intervalusing
usingthe
thedata
datain
inExample
Example8-6:
8-6:

 1  1
( x1  x 2 )  t
2
sp    ( 6870  6598 )  2 .06 ( 595835)( 0.15)  [ 343.85,887 .85]
  n1 n2 
2
 8-27

8-4 A Large-Sample Test for the Difference

between Two Population Proportions

• Hypothesized difference is zero

 I: Difference between two population proportions is 0
• p1 = p2
» H0: p1 -p2 = 0
» H1: p1 -p20
 II: Difference between two population proportions is less than 0
• p1p2
» H0: p1 -p2  0
» H1: p1 -p2 > 0
• Hypothesized difference is other than zero:
 III: Difference between two population proportions is less than D
• p1 p2+D
» H0:p-p2  D
» H1: p1 -p2 > D
 8-28

Comparisons of Two Population Proportions When the

Hypothesized Difference Is Zero: Test Statistic

Whenthe
When thepopulation
populationproportions
proportionsare
arehypothesized
hypothesizedtotobebeequal,
equal,then
thenaapooled
pooledestimator
estimatorofof
theproportion
the proportion(( p ))may
maybebeused
usedinincalculating
calculatingthe
thetest
teststatistic.
statistic.

AAlarge-sample
large-sampletest
teststatistic
statisticfor
forthe
thedifference
differencebetween
betweentwo twopopulation
population
proportions,when
proportions, whenthethehypothesized
hypothesizeddifference
differenceisiszero:
zero:

( pˆ 1  pˆ 2 )  0
z
1 1
pˆ (1  pˆ )  
 n1 n2 

where
where pˆ 1 
x1
isisthe
thesample
sampleproportion
proportionininsample and p  x1 isisthe
sample 11and thesample
sample
n1 1
n1
proportionininsample
proportion sample2.2. The symbol p stands
Thesymbol standsfor
forthe
thecombined
combinedsample
sample
proportionininboth
proportion bothsamples,
samples,considered
consideredas
asaasingle
singlesample.
sample. That
Thatis:
is:

x1  x1
pˆ 
n1  n2
 8-29

Comparisons of Two Population Proportions When the

Hypothesized Difference Is Zero: Example 8-7

Carry out a two-tailed test of the equality of banks’ share of the car loan market in 1980 and 1995.

Population 1:1980
H : p  p 0
n = 100 0 1 2
1 H : p  p 0
x = 53 1 1 2
1 ( p  p )0
1 2 0.53  0.43
z 
p̂ = 0.53    1 1 
1 pˆ (1 pˆ ) 1  1  (.48)(.52)  
n n   100 100 
 1 2
0.10 0.10
Population 2 :1995    1.415
0.004992 0.07065
n 2 = 100
x 2 = 43 Critical point : z = 1.645
0.05
p̂2 = 0.43 H may not be rejected even at a 10%
0
x +x level of significance.
pˆ  1 2  53  43  0.48
n  n 100 100
1 2
 8-30

Example 8-7: Carrying Out the Test

Standard Normal Distribution Sincethe

Since thevalue
valueofofthe
thetest
test
0.4
statisticisiswithin
statistic withinthe
the
0.3
nonrejectionregion,
nonrejection region,even
evenatataa
10%level
levelof ofsignificance,
significance,we we
f(z)

0.2
10%
0.1
mayconclude
may concludethat thatthere
thereisisno
no
0.0
0
z statisticallysignificant
statistically significant
-z0.05=-1.645 z0.05=1.645
Rejection Nonrejection Rejection differencebetween
difference betweenbanks’
banks’
Region Region Region
Test Statistic=1.415 sharesof
shares ofcarcarloans
loansinin1980
1980
and1995.
and 1995.
 8-31

Comparisons of Two Population Proportions When

the Hypothesized Difference Is Not Zero: Example 8-8

Carryout
Carry outaaone-tailed
one-tailedtest
testtotodetermine
determinewhether
whetherthe
thepopulation
populationproportion
proportionofoftraveler’s
traveler’scheck
checkbuyers
buyerswho
whobuybuy
atatleast
least$2500
$2500ininchecks
checkswhen
whensweepstakes
sweepstakesprizes
prizesare
areoffered
offeredasasatatleast
least10%
10%higher
higherthan
thanthe
theproportion
proportionofofsuch
such
buyerswhen
buyers whennonosweepstakes
sweepstakesare areon.
on.

Population 1: With Sweepstakes

n = 300 H : p  p  0.10
1 0 1 2
H : p  p  0.10
x = 120 1 1 2
1 ( pˆ  pˆ )  D
p̂ = 0.40 z 1 2
1
 1(1 p1) p2 (1 p2 ) 
pˆ ˆ ˆ ˆ 
  
 n n 
 1 2 

 (0.40  0.20)  0.10 

0.10
 3.118
Population 2 : No Sweepstakes  (0.40)(0.60) (0.20)(.80)  0.03207
  
 300 700 
n 2 = 700 

x 2 = 140 Critical point : z = 3.09

0.001
p̂2 = 0.20 H may be rejected at any common level of significance.
0
 8-32

Example 8-8: Carrying Out the Test

Standard Normal Distribution Sincethe

Since thevalue
valueof ofthe
thetest
test
0.4
statisticisisabove
statistic abovethethecritical
critical
0.3
point,even
point, evenfor foraalevel
levelof
of
significanceas
significance assmall
smallasas0.001,
0.001,
f(z)

0.2

0.1
thenull
the nullhypothesis
hypothesismay maybebe
0.0
0
z rejected,and
rejected, andwe wemay
may
z0.001=3.09

Nonrejection Rejection concludethat

conclude thatthe
theproportion
proportion
Region
Region
Test Statistic=3.118 ofcustomers
of customersbuyingbuyingatatleast
least
$2500of
$2500 oftravelers
travelerschecks
checksisisatat
least10%
least 10%higher
higherwhen
when
sweepstakesare
sweepstakes areon.
on.
 8-33

Confidence Intervals for the Difference

between Two Population Proportions
AA(1-)
(1-)100%
100%large-sample
large-sampleconfidence
confidenceinterval
intervalfor
forthe
thedifference
difference
betweentwo
between twopopulation
populationproportions:
proportions:


 pˆ (1 pˆ ) pˆ (1 pˆ ) 
( pˆ  pˆ )  z 1 1  2 2
1 2  n n 
 
 1 2 
2

AA95%
95%confidence
confidenceinterval
intervalusing
usingthe
thedata
datain
inexample
example8-8:
8-8:

 pˆ (1 pˆ ) pˆ (1 pˆ ) 
( pˆ  pˆ )  z 1 1  2 2  (0.4  0.2) 1.96 (0.4)(0.6)  (0.2)(0.8)
1 2  n n 

 1 2 
 300 700
2
 0.2  (1.96)(0.0321)  0.2  0.063  [0.137,0.263]
 8-34

8-5 The F Distribution and a Test for

Equality of Two Population Variances

TheFFdistribution
The distributionisisthe
thedistribution
distributionofofthe
theratio
ratioofoftwo
twochi-square
chi-squarerandom
randomvariables
variables
thatare
that areindependent
independentofofeach
eachother,
other,each
eachofofwhich
whichisisdivided
dividedbybyits
itsown
owndegrees
degreesofof
freedom.
freedom.

AnFFrandom
An randomvariable
variablewith
withkk1and
andkk2degrees
degreesofoffreedom:
freedom:
1 2

 12
k1
F k ,k  
1 2
 22
k2
 8-35

The F Distribution

•• The
TheFFrandom
randomvariable
variablecannot
cannot F Distributions with different Degrees of Freedom
benegative,
be negative,so soititisisbound
boundbyby
zeroon
zero onthe
theleft.
left. 1.0 F(25,30)

f(F)
•• The
TheFFdistribution
distributionisisskewed
skewedtoto F(10,15)

theright.
the right. 0.5

•• The
TheFFdistribution
distributionisisidentified
identified
thenumber
the numberof ofdegrees
degreesof of 0.0
F(5,6)

freedomin
freedom inthe
thenumerator,
numerator,kk,1, 0 1 2 3 4 5
F
1
andthe
and thenumber
numberof ofdegrees
degreesofof
freedomin
freedom inthe
thedenominator,
denominator,
kk2.2.
 8-36

Using the Table of the F Distribution

Critical Points of the F Distribution Cutting Off a F Distribution with 7 and 11 Degrees of Freedom
Right-Tail Area of 0.05

k1 1 2 3 4 5 6 7 8 9 0.7

0.6
k2
1 161.4 199.5 215.7 224.6 230.2 234.0 236.8 238.9 240.5 0.5
2 18.51 19.00 19.16 19.25 19.30 19.33 19.35 19.37 19.38 0.4

f(F)
3 10.13 9.55 9.28 9.12 9.01 8.94 8.89 8.85 8.81
4 7.71 6.94 6.59 6.39 6.26 6.16 6.09 6.04 6.00 0.3
5 6.61 5.79 5.41 5.19 5.05 4.95 4.88 4.82 4.77 0.2
6 5.99 5.14 4.76 4.53 4.39 4.28 4.21 4.15 4.10
7 5.59 4.74 4.35 4.12 3.97 3.87 3.79 3.73 3.68 0.1
8 5.32 4.46 4.07 3.84 3.69 3.58 3.50 3.44 3.39 0.0 F
9 5.12 4.26 3.86 3.63 3.48 3.37 3.29 3.23 3.18 0 1 2 3 4 5
10 4.96 4.10 3.71 3.48 3.33 3.22 3.14 3.07 3.02
11 4.84 3.98 3.59 3.36 3.20 3.09 3.01 3.01 2.95 2.90
12 4.75 3.89 3.49 3.26 3.11 3.00 2.91 2.85 2.80 F0.05=3.01
13 4.67 3.81 3.41 3.18 3.03 2.92 2.83 2.77 2.71
14 4.60 3.74 3.34 3.11 2.96 2.85 2.76 2.70 2.65
15 4.54 3.68 3.29 3.06 2.90 2.79 2.71 2.64 2.59

The left-hand critical point to go along with F(k1,k2) is given by: 1

F k 2 ,k 1
Where F(k1,k2) is the right-hand critical point for an F random variable with the reverse
number of degrees of freedom.
 8-37

Critical Points of the F Distribution:

F(6, 9),  = 0.10

F Distribution with 6 and 9 Degrees of Freedom Theright-hand

The right-handcritical
criticalpoint
pointread
read
directlyfrom
directly fromthethetable
tableofofthe
theFF
0.7
distributionis:
distribution is:
0.05
0.6 0.90
0.5

0.4
FF(6,9) =3.37
(6,9)=3.37
f(F)

0.3
0.05
0.2

0.1 Thecorresponding
The correspondingleft-hand
left-handcritical
critical
0.0 pointisisgiven
point givenby:
by:
0 1 2 3 4 5 F
1 1
F0.95=(1/4.10)=0.2439 F0.05=3.37
  0.2439
F 9 , 6 410
.
 8-38

Test Statistic for the Equality of Two

Population Variances
Teststatistic
Test statisticfor
forthe
theequality
equalityof
ofthe
thevariances
variancesof
oftwo
twonormally
normally
distributedpopulations:
distributed populations:
2
s
s1 2

FF n n1,1n,n11  s122

s2
1 2
1 2

 I:I:Two-Tailed
Two-TailedTest
Test
•• 1==2 1 2

•• HH[Link]1 1==2 2
•• HH:
1:2
1  2

 II:
II:One-Tailed
One-TailedTest
Test
•• 
12
1 2

•• HH[Link]1 1
2
2

•• HH:1:1
1 1
2
2
 8-39

Example 8-9
The economist wants to test whether or not the event (interceptions and prosecution of insider
traders) has decreased the variance of prices of stocks.

Population 1 : Before
n = 25 2
1 2 2
H 0:  
s 2  9.3 1 21
1
2 2
Population 2 : After H1:  
n = 24 1 2
2
s2
s 2  3.0 1 9.3
F  F    3.1
2
n1  1, n 2  1 
24,23 s2  3.0
2
  0.05
F  2.01
24,23 H 0 may be rejected at a 1% level of significance.
  0.01
F  2.70
24,23
 8-40

Example 8-9: Solution

Distribution with 24 and 23 Degrees of Freedom Sincethe

Since thevalue
valueofofthe
thetest
test
0.7

0.6
statisticisisabove
statistic abovethethecritical
critical
0.5 point,even
point, evenforforaalevel
levelof
of
0.4
significanceas
significance assmall
smallasas0.01,
0.01,
f(F)

0.3

0.2 thenull
the nullhypothesis
hypothesismay maybebe
0.1
rejected,and
rejected, andwewemaymay
0.0 F
0 1 2 3 4 5 concludethat
conclude thatthe
thevariance
varianceofof
F0.01=2.7 Test Statistic=3.1 stockprices
stock pricesisisreduced
reducedafter
after
theinterception
the interceptionand and
prosecutionof
prosecution ofinside
insidetraders.
traders.
 8-41

Example 8-10: Testing the Equality of

Variances for Example 8-5

Population 1 Population 2
n = 14 n =9 2
H :  
2
1 2 0 1 2
2 2 2 2
s  0.12 s  0.11 2 2
1 2 H :  
1 1 2
  0.05
s2
F  3.28 1 0.122
F F    119
.
13,8  n1  1, n 2  1   s 0.11
13,8 2 2
2
  0.10
F  2.50 H may not be rejected at the 10% level of significance.
13,8 0
 8-42

Example 8-10: Solution

F Distribution with 13 and 8 Degrees of Freedom

Sincethe
Since thevalue
valueof ofthe
thetest
test
statisticisisbetween
betweenthe thecritical
critical
0.7

0.10
0.6 0.80 statistic
0.5

0.4
points,even
points, evenfor foraa20%
20%level
levelof
of
f(F)

0.3
0.10 significance,we
significance, wecan
cannot
notreject
reject
0.2

0.1 thenull
the nullhypothesis.
hypothesis. We We
0.0
0 1 2 3 4 5 F concludethe
conclude thetwo
twopopulation
population
F0.90=(1/2.20)=0.4545 F0.10=3.28
variancesare
variances areequal.
equal.
Test Statistic=1.19
 8-43

Test for the Difference between Two

Population Variances: Example 8-10

The p-value
The p-value forfor the
the test
test is
is 0.8304
0.8304
which is
which is larger
larger than
than 0.05.
0.05. Thus
Thus
the null
the null hypothesis
hypothesis cannot
cannot be be
rejected at
rejected at this
this level
level of
of
significance of
significance of 0.05.
0.05. That
That is,is, one
one
can assume
can assume equal
equal variance.
variance.