6-1

Estimation
 6-2

Types of estimators
• Point Estimate
A single-valued estimate.
A single element chosen from a sampling
distribution.
Conveys little information about the actual value of
the population parameter, about the accuracy of
the estimate.
• Confidence Interval or Interval Estimate
An interval or range of values believed to include
the unknown population parameter.
Associated with the interval is a measure of the
confidence we have that the interval does indeed
contain the parameter of interest.
 6-3

1. Point Estimation
Definition:
A parameter is a numerical descriptive measure

of a population ( μ is an example of a parameter).

A statistic is a numerical descriptive measure of

a sample ( X is an example of a statistic).

To each sample statistic there corresponds a

population parameter.
We use X , S2, S , p, etc. to estimate μ, σ2, σ, P (or

π), etc.
 6-4

Sample statistics Population

ˉ (sample mean)
X
parameter
S2 ( sample  μ (population
variance) mean)
S (sample Standard σ2 ( population
deviation) variance)
P (sample σ(population
proportion) standard deviation)
P or π (Population
proportion)
 6-5

Definition:
A point estimate of some population parameter
O is a single value Ô of a sample statistic

Sampling Distribution of Means

one of the most fundamental concepts of
statistical inference, and it has remarkable
properties.
Since it is a frequency distribution it has its own
mean and standard deviation
we shall use the notation for the standard
deviation of the distribution.
The standard deviation of the sampling
distribution of means is called the standard
error of the mean.
 6-6

Properties
1. The mean of the sampling distribution of
means is the same as the population mean, μ .

2. The SD of the sampling distribution of means

is σ / √n .

3. The shape of the sampling distribution of

means is approximately a normal curve,
regardless of the shape of the population
distribution and provided n is large enough
(Central limit theorem).
 6-7

Confidence Intervals

Using Statistics

Confidence Interval for the Population Mean
When the Population Standard Deviation is
Known

Confidence Intervals for  When  is Unknown -
The t Distribution

Large-Sample Confidence Intervals for the
Population Proportion p

Sample Size Determination
 6-8

• Consider the following statements:

x = 550
• A single-valued estimate that conveys little information
about the actual value of the population mean.
We are 99% confident that  is in the interval [449,551]
• An interval estimate which locates the population mean
within a narrow interval, with a high level of confidence.
We are 90% confident that  is in the interval [400,700]
• An interval estimate which locates the population mean
within a broader interval, with a lower level of confidence.
 6-9

AAconfidence
confidenceinterval
intervalor orinterval
intervalestimate
estimateisisaarange
rangeor
orinterval
intervalof
of
numbersbelieved
numbers believedtotoinclude
includeananunknown
unknownpopulation
populationparameter.
parameter.
Associatedwith
Associated withthe
theinterval
intervalisisaameasure
measureofofthe
theconfidence
confidencewe wehave
have
thatthe
that theinterval
intervaldoes
doesindeed
indeedcontain
containthe
theparameter
parameterof ofinterest.
interest.

•A confidence interval or interval

estimate has two components:
A range or interval of values
An associated level of confidence
 6-10


If the population distribution is normal, the
sampling distribution of the mean is normal.
If the sample is sufficiently large, regardless of
the shape of the population distribution, the
sampling distribution is normal (Central Limit
Theorem).

In either case : Standard Normal Distribution: 95% Interval

In either case :
0.4
   
    1.96  x    1.96   0.95
P
P   1.96 n x    1.96 n 0.95
 n n 0.3

f(z)
or 0.2
or

0.1
   
  x  1.96    x  1.96   0.95
P
P x  1.96 n   x  1.96 n 0.95
 n n 0.0
-4 -3 -2 -1 0 1 2 3 4
z
 6-11

Aftersampling,
After sampling,approximat
approximately
ely95%
95%of
of such
suchintervals
intervals


x 1.96
x 1.96
nn
willinclude
will includethe
thepopulation
populationmean
mean(and
(and5%
5%of
of them
themwill
willnot).
not).



Thatis,
That 1.96
is,xx 1.96 isisaa95%
95%confidence
confidenceinterval for..
intervalfor
nn
 6-12

Approximately95%
Approximately 95%of ofthe
theintervals
intervals
Sampling
p n Distribution
u of the Mean aroundthe
x 1.96  around thesample
samplemean
meancan
canbebe
n
0.4
expectedtotoinclude
expected includethe
theactual
actualvalue
valueof
ofthe
the
0.3
95%
populationmean,
population mean,.. (When
(Whenthethesample
sample
meanfalls
mean fallswithin
withinthe
the95%
95%interval
intervalaround
around
f(x)

0.2

0.1
thepopulation
the populationmean.)
mean.)
2.5% 2.5%
x x x

**5%
0.0

  196
.

n
  196
.

n
x
5%of
ofsuch
suchintervals
intervalsaround
aroundthe
thesample
sample
meancan
mean canbebeexpected
expectednot
nottotoinclude
includethe
the
x
actualvalue
actual valueof
ofthe
thepopulation
populationmean.
mean.
x
(Whenthe
(When thesample
samplemean
meanfalls
fallsoutside
outsidethe
the
x
95%interval
intervalaround
aroundthe
thepopulation
population
* x
x
95%
mean.)
x mean.)
x

x
x
x
x
*
 6-13

Interpretation:
a.Probabilistic: in repeated sampling, 100(1-α)% of all
intervals will include μ
b.Practical: we are 100(1-α)% confident that a interval
contains μ.
 6-14

AA95%
95%confidence
confidenceinterval forwhen
intervalfor whenisisknown
knownandandsampling
samplingisis
donefrom
done fromaanormal
normalpopulation,
population,ororaalarge
largesample
sampleisisused:
used:

x 1.96
n

Thequantity
quantity 1.96 
The isisoften
oftencalled
calledthe
themargin
marginof
oferror
erroror
orthe
the
n
samplingerror.
sampling error.

For example, if: n = 25 A 95% confidence interval:

= 20
x = 122 x 1.96

n
122 1.96
20
25
122 (1.96)(4)
122 7.84
114 .16,129.84
 6-15

z 
2 2

(1   )

 
2 2

 z z
2
2
 6-16

(1   )
 z
2 2 (1   )

0.99 0.005 2.576

0.98 0.010 2.326  
2 2
0.95 0.025 1.960
0.90 0.050 1.645  z
2
z
2

0.80 0.100 1.282

 6-17

Whensampling
When samplingfrom
fromthe
thesame
samepopulation,
population,using
usingaafixed
fixedsample
samplesize,
size,the
the
higherthe
higher theconfidence
confidencelevel,
level,the
thewider
widerthe
theconfidence
confidenceinterval.
interval.
St an d ar d N or m al Dis tri b uti o n St an d ar d N or m al Di s tri b uti o n

0.4 0.4

0.3 0.3
f(z)

f(z)
0.2 0.2

0.1 0.1

0.0 0.0
-5 -4 -3 -2 -1 0 1 2 3 4 5 -5 -4 -3 -2 -1 0 1 2 3 4 5
Z Z

80% Confidence Interval: 95% Confidence Interval:

 
x 128
. x 196
.
n n
 6-18

Whensampling
When samplingfrom
fromthe
thesame
samepopulation,
population,using
usingaafixed
fixedconfidence
confidence
level,the
level, thelarger
largerthe
thesample
samplesize,
size,n,
n,the
thenarrower
narrowerthetheconfidence
confidence
interval.
interval.
S am
S am pplin
lingg D
D is
istrib
tributio
utionn ooff th
thee M
Meean
an S am
S am pplin
lingg D
D is
istrib
trib utio
utio nn ooff th
thee M
Mee an
an

00.4
.4 00.9
.9

00.8
.8

00.3
.3 00.7
.7

00.6
.6

00.5
.5

f(x)
f(x)

00.2
.2

f(x)
f(x)

00.4
.4

00.3
.3
00.1
.1
00.2
.2

00.1
.1

00.0
.0
00.0
.0

xx xx

95% Confidence Interval: n = 20 95% Confidence Interval: n = 40

 6-19

A
 A physical
physical therapist
therapist wished
wished to
to estimate,
estimate, with
with
99% confidence,
99% confidence, the
the mean
mean maximal
maximal strength
strength
of aa particular
of particular muscle
muscle in in aa certain
certain group
group of
of
individuals. He
individuals. He assume
assume thatthat strength
strength scores
scores
are approximately
are approximately normally
normally distributed
distributed with
with
aa variance
variance of of 144.
144. AA sample
sample of of 15
15 subjects
subjects
who participated
who participated inin the
the experiment
experiment yielded
yielded aa
mean of
mean of 84.3.
84.3. What
What is is 90%
90% CI?CI?
 6-20

Solution
α = 0.01⇒ Zα/2 = 2.58
Mean =84.3, n=15, σ =12
84.3 ± 2.58(12/ √15) ⇒ 84.3 ± 8.0 ⇒ (76.3, 92.3)
⇒ We are 99% confident that the population mean is
between 76.3 and 92.3.
 6-21

If the population standard deviation, , is not known, replace

with the sample standard deviation, s. If the population is
normal, the resulting statistic: t  X s 
n
has a t distribution with (n - 1) degrees of freedom.
•• Thet tisisaafamily
The familyofofbell-shaped
bell-shapedand andsymmetric
symmetric
Standard normal
distributions,one
distributions, oneforforeach
eachnumber
numberofofdegree
degreeofof
freedom.
freedom. t, df = 20
•• Theexpected
The expectedvalue
valueofoft tisis0.0.
t, df = 10
•• Fordfdf>>2,2,the
For thevariance
varianceofof t tisisdf/(df-2).
df/(df-2). This
Thisisis
greaterthan
greater than1,1,but
butapproaches
approaches11asasthe thenumber
number
ofofdegrees
degreesofoffreedom
freedomincreases.
increases. The Thet tisisflatter
flatter
andhas
and hasfatter
fattertails
tailsthan
thandoes
doesthe thestandard
standard
normal.
normal. 
•• Thet tdistribution
The distributionapproaches
approachesaastandard
standardnormal
normal 

asasthe
thenumber
numberofofdegrees
degreesofoffreedom
freedomincreases
increases
 6-22

(1-)100%
AA(1- )100%confidence
confidenceinterval forwhen
intervalfor whenisisnot
notknown
known
(assumingaanormally
(assuming normallydistributed
distributedpopulation):
population):
s
x t
n

2

wheret isisthe
where thevalue
valueofofthe
thettdistribution
distributionwith
withn-1n-1degrees
degreesof
of
2 
freedomthat
freedom thatcuts
cutsoff
offaatail
tailarea
areaof
of 2 totoits
itsright.
right.
 6-23

df t0.100 t0.050 t0.025 t0.010 t0.005

--- ----- ----- ------ ------ ------
1 3.078 6.314 12.706 31.821 63.657 t D is trib utio n: d f = 1 0
2 1.886 2.920 4.303 6.965 9.925
3 1.638 2.353 3.182 4.541 5.841 0 .4
4 1.533 2.132 2.776 3.747 4.604
5 1.476 2.015 2.571 3.365 4.032
6 1.440 1.943 2.447 3.143 3.707 0 .3

7 1.415 1.895 2.365 2.998 3.499 Area = 0.10 Area = 0.10

8 1.397 1.860 2.306 2.896 3.355

}
f(t)
0 .2
9 1.383 1.833 2.262 2.821 3.250
10 1.372 1.812 2.228 2.764 3.169
11 1.363 1.796 2.201 2.718 3.106 0 .1
12 1.356 1.782 2.179 2.681 3.055
13 1.350 1.771 2.160 2.650 3.012
14 1.345 1.761 2.145 2.624 2.977 0 .0
15 1.341 1.753 2.131 2.602 2.947 -1.372 0 1.372
-2.228 2.228
16 1.337 1.746 2.120 2.583 2.921

}
t
17 1.333 1.740 2.110 2.567 2.898
18 1.330 1.734 2.101 2.552 2.878
19 1.328 1.729 2.093 2.539 2.861 Area = 0.025 Area = 0.025
20 1.325 1.725 2.086 2.528 2.845
21 1.323 1.721 2.080 2.518 2.831
22
23
1.321
1.319
1.717
1.714
2.074
2.069
2.508
2.500
2.819
2.807 Wheneverisisnot
Whenever notknown
known(and
(andthe
thepopulation
populationisis
24
25
1.318
1.316
1.711
1.708
2.064
2.060
2.492
2.485
2.797
2.787 assumednormal),
assumed normal),thethecorrect
correctdistribution
distributiontotouse
useisis
26
27
1.315
1.314
1.706
1.703
2.056
2.052
2.479
2.473
2.779
2.771 thet tdistribution
the distributionwith
withn-1
n-1degrees
degreesofoffreedom.
freedom.
28
29
1.313
1.311
1.701
1.699
2.048
2.045
2.467
2.462
2.763
2.756 Note,however,
Note, however,that
thatfor
forlarge
largedegrees
degreesofoffreedom,
freedom,
30
40
1.310
1.303
1.697
1.684
2.042
2.021
2.457
2.423
2.750
2.704 thet tdistribution
the distributionisisapproximated
approximatedwellwellbybythe
theZZ
60
120
1.296
1.289
1.671
1.658
2.000
1.980
2.390
2.358
2.660
2.617 distribution.
distribution.
 1.282 1.645 1.960 2.326 2.576
 6-24

A study of hypoxemia during the immediate post-operative period reported

the fractions of ideal weight for 11 patients who became severely hypoxemic
during transfer to the recovery room. The mean is 1.51 and the standard
deviation is 0.33. Estimate the 95% C.I. for the population mean fraction of
ideal weight, where the population consists of hypoxemic patients similar to
those in the study (The data is normally distributed, use α=0.05).

Solution
• tα/2, n-1 / = t 0.025,10 = 2.2281
1 . 51 ± 2 . 2281(0 . 33/11)
1 . 51± 0 . 221
(1 . 289 ,1 . 731 )
• We are 95% sure that the μ (1 . 289 ,1 . 731 ) population mean lies
between 1.289 and 1.731
 6-25

df t0.100 t0.050 t0.025 t0.010 t0.005

--- ----- ----- ------ ------ ------ Whenever  is not known (and the population is
1 3.078 6.314 12.706 31.821 63.657
. . . . . . assumed normal), the correct distribution to use is
. . . . . .
. . . . . . the t distribution with n-1 degrees of freedom.
120 1.289 1.658 1.980 2.358 2.617
 1.282 1.645 1.960 2.326 2.576 Note, however, that for large degrees of freedom,
the t distribution is approximated well by the Z
distribution.
 6-26

A large - sample (1 -  )100% confidence interval for  :

s
x z 
2 n

Example 6-3: An economist wants to estimate the average amount in checking accounts
at banks in a given region. A random sample of 100 accounts gives x-bar = $357.60
and s = $140.00. Give a 95% confidence interval for , the average amount in any
checking account at a bank in the given region.

s 140.00
x z0.025 357.60 196
. 357.60 27.44  33016,38504
. . 
n 100
 6-27

The estimator of the population proportion, p , is the sample proportion, p . If the

sample size is large, p has an approximately normal distribution, with E( p ) = p and
pq
V( p ) = , where q = (1 - p). When the population proportion is unknown, use the
n
estimated value, p , to estimate the standard deviation of p .

For estimating p , a sample is considered large enough when both n p an n q are greater
than 5.
 6-28

A large - sample (1 -  )100% confidence interval for the population proportion, p :

pˆ qˆ
pˆ z
α n
2
where the sample proportion, p̂, is equal to the number of successes in the sample, x,
divided by the number of trials (the sample size), n, and q̂ = 1 - p̂.
 6-29

AAmarketing
marketingresearch
researchfirm
firmwants
wantstotoestimate
estimatethe
theshare
sharethat
thatforeign
foreigncompanies
companies
haveininthe
have theAmerican
Americanmarket
marketfor forcertain
certainproducts.
products. AArandom
randomsample
sampleof
of100
100
consumersisisobtained,
consumers obtained,and
andititisisfound
foundthat
that34
34people
peopleininthe
thesample
sampleare
areusers
users
offoreign-made
of foreign-madeproducts;
products;the
therestrestare
areusers
usersof
ofdomestic
domesticproducts.
products. Give
Giveaa
95%confidence
95% confidenceinterval
intervalfor
forthe
theshare
shareof
offoreign
foreignproducts
productsininthis
thismarket.
market.


pq ( 0.34 )( 0.66)
p z  0.34 1.96
2
n 100
 0.34 (1.96)( 0.04737 )
 0.34 0.0928
 0.2472 ,0.4328

Thus,the
Thus, thefirm
firmmay
maybebe95%
95%confident
confidentthat
thatforeign
foreignmanufacturers
manufacturerscontrol
control
anywherefrom
anywhere from24.72%
24.72%toto43.28%
43.28%ofofthe
themarket.
market.
 6-30

The width of a confidence interval can be reduced only at the

price of:
• a lower level of confidence, or
• a larger sample.

LowerLevel
Lower Levelof
ofConfidence
Confidence LargerSample
Larger SampleSize
Size

90% Confidence Interval Sample Size, n = 200


pq (0.34)(0.66) 
pq (0.34)(0.66)
p z  0.34 1645
. p z  0.34 196
.
2
n 100 2
n 200
0.34 (1645
. )(0.04737)  0.34 (196
. )(0.03350)
0.34 0.07792  0.34 0.0657
 0.2621,0.4197  0.2743,0.4057
 6-31

Beforedetermining
Before determiningthe
thenecessary
necessarysample
samplesize,
size,three
threequestions
questionsmust
must
beanswered:
be answered:

•• How
Howclose
closedo
doyou
youwant
wantyour
yoursample
sampleestimate
estimateto
tobe
beto
tothe
the
unknownparameter?
unknown parameter? (What
(Whatisisthe
thedesired bound, B?)
desired bound, B?)
•• What
Whatdo doyou
youwant
wantthe
thedesired
desiredconfidence
confidencelevel (1-)to
level (1-) to
beso
be sothat
thatthe
thedistance
distancebetween
betweenyour
yourestimate
estimateand
andthe
the
parameteris
parameter isless
lessthan
thanor
orequal
equalto
toB?
B?
•• What
Whatis isyour
yourestimate
estimateof
ofthe
thevariance
variance(or
(orstandard
standard
deviation)of
deviation) ofthe
thepopulation
populationin
inquestion?
question?

For example : A (1 -  ) Confidence Interval for  : x z 

For example : A (1 -  ) Confidence Interval for  : x z 
22 n
n
 6-32

Thesample
The samplesize
sizedetermines
determinesthe
thebound
boundofofaastatistic,
statistic,since
sincethe
thestandard
standard
errorof
error ofaastatistic
statisticshrinks
shrinksas
asthe
thesample
samplesize
sizeincreases:
increases:

Sample size = 2n
Standard error
of statistic

Sample size = n
Standard error
of statistic


6-33
 6-34

AAmarketing
marketingresearch
researchfirm
firmwants
wantstotoconduct
conductaasurvey
surveytotoestimate
estimatethe
theaverage
average
amountspent
amount spentononentertainment
entertainmentby byeach
eachperson
personvisiting
visitingaapopular
popularresort.
resort. The
The
peoplewho
people whoplan
planthe
thesurvey
surveywould
wouldlike
liketotodetermine
determinethe
theaverage
averageamount
amountspent
spentby
by
allpeople
all peoplevisiting
visitingthe
theresort
resorttotowithin
within$120,
$120,with
with95%
95%confidence.
confidence. From
Frompast
past
operationof
operation ofthe
theresort,
resort,ananestimate
estimateofofthe
thepopulation
populationstandard
standarddeviation
deviationisis
ss==$400.
$400. What
Whatisisthe
theminimum
minimumrequired
requiredsample
samplesize?
size?


zz 2
2
2
2

n
n  2
2

BB 2
2

((1
1..96
96)) ((400
400))
2 2
2 2

 120
120 2
2

42
 42..684 43
684  43
 6-35

Themanufacturers
The manufacturersof ofaasports
sportscar
carwant
wanttotoestimate
estimatethe
theproportion
proportionof
ofpeople
peopleininaa
givenincome
given incomebracket
bracketwho
whoare
areinterested
interestedininthe
themodel.
model. The
Thecompany
companywants
wantstoto
knowthe
know thepopulation
populationproportion,
proportion,p,p,totowithin
within0.01
0.01with
with99%
99%confidence.
confidence. Current
Current
companyrecords
company recordsindicate
indicatethat
thatthe
theproportion
proportionppmaymaybebearound
around0.25.
0.25. What
Whatisisthe
the
minimumrequired
minimum requiredsample
samplesize
sizefor
forthis
thissurvey?
survey?
2
z2 pq
zpq
n 
n  B22
2
2

B
2
2 .576
2 ( 0.25)( 0.75)
2.576 (0.25)(0.75)

 2
010
010
. . 2

124.42
 125
124.42 125