POINT ESTIMATION

Note: The Slides were taken from Elementary

Statistics: A Handbook of Slide Presentation
prepared by Z.V.J. Albacea, C.E. Reano, R.V.
Collado, L.N. Comia and N.A. Tandang in 2005
for the Institute of Statistics, CAS, UP Los
Banos
 TEACHING BASIC STATISTICS …

PROBLEMS ADDRESSED BY
INFERENTIAL STATISTICS

ESTIMATION is concerned with

finding a value or range of values for
an unknown parameter.
TEST OF HYPOTHESIS deals with
evaluating a claim or a conjecture
about a parameter or distribution of
the population.

Session 5.2
 TEACHING BASIC STATISTICS …

Basic Idea of Estimation

Population Random Sample

I am 95%
Mean confident that
Mean, m, is m is between
unknown X = 50
40 & 60.

Sample

Session 5.3
 TEACHING BASIC STATISTICS …

ESTIMATION

How do we estimate a parameter, q?

Specifically, how do we estimate
▪ a population mean, m ?
▪ a population standard deviation, s?
▪ a population proportion, p?

Session 5.4
 TEACHING BASIC STATISTICS …

ESTIMATION

▪ An estimator of a parameter is a
rule or a formula for computing
an estimate using the sample
data.
▪ It is usually denoted by a Greek
letter with a ‘hat’ like qˆ and m̂ .

Session 5.5
 TEACHING BASIC STATISTICS …

ESTIMATION

▪ In other cases, special symbols

are used like X for the sample
mean as estimator of the
population mean.
▪ An estimate is a numerical value
of the estimator.

Session 5.6
 TEACHING BASIC STATISTICS …

ESTIMATION

There can be several estimators for a

particular parameter.
For example, a population mean can
be estimated by any one of the
following:
▪sample mean
▪sample modal value
▪sample median
Session 5.7
 TEACHING BASIC STATISTICS …

GOOD PROPERTY OF AN ESTIMATOR

BIAS of an estimator qˆ is given by:

( )
BIAS qˆ,q = E qˆ  − q

where E qˆ  is the expected value of the estimator

and q is the parameter of interest.

An estimator with its bias equal to zero is said to be

an unbiased estimator of q. Unbiasedness is a
desirable property of an estimator.

Session 5.8
 TEACHING BASIC STATISTICS …

GOOD PROPERTY OF AN ESTIMATOR

An estimator must be
ACCURATE.
Accuracy measures
the closeness of an
estimate to its true
value.

Session 5.9
 TEACHING BASIC STATISTICS …

MEASURE OF ACCURACY

▪ To measure accuracy, we use the Mean

Square Error or MSE.

( ) ( )
2
MSE qˆ = E qˆ − q
( )
= BIAS qˆ,q + VAR qˆ
2
( )
▪ For an unbiased estimator, its MSE is
equal to its Variance.
▪ An accurate estimator is one whose MSE
is small.

Session 5.10
 TEACHING BASIC STATISTICS …

GOOD PROPERTY OF AN ESTIMATOR

An estimator must be
PRECISE.
Precision measures the
closeness of the different
possible values of the
estimator to each other.

Session 5.11
 TEACHING BASIC STATISTICS …

MEASURE OF PRECISION
The precision of an estimator can be measured by
its variance or by its standard error which is the
square root of the variance.
We can estimate the standard error using sample
data. For example, if we have a random sample
then we can estimate the standard error of the
sample mean:
s
sˆ X =
n

Session 5.12
 TEACHING BASIC STATISTICS …

ACCURACY VS. PRECISION

This estimator is not accurate. It is not unbiased
and not precise.
*
**
**** *
* *
q *

Session 5.13
 TEACHING BASIC STATISTICS …

ACCURACY VS. PRECISION

This estimator is accurate. It is both unbiased and

precise.

* **q***
****
*

Session 5.14
 TEACHING BASIC STATISTICS …

ACCURACY VS. PRECISION

This estimator is not accurate. It is

unbiased but not precise.

**
** q* *
* *
* * *

Session 5.15
 TEACHING BASIC STATISTICS …

ACCURACY VS. PRECISION

This estimator is not accurate. It is precise
but not unbiased.

*********
**
q

Session 5.16
 TEACHING BASIC STATISTICS …

IN CHOOSING AN ESTIMATOR…

Would you choose

• an unbiased estimator over a less
precise estimator?
• a biased estimator over a more precise
estimator?
• an estimator which is both precise and
unbiased?

Session 5.18
 TEACHING BASIC STATISTICS …

TWO TYPES OF ESTIMATOR

1. A POINT ESTIMATOR is a formula that

gives a single value in estimating a
parameter.
EXAMPLE: X is a point estimator of m
s is a point estimator of s
p̂ is a point estimator of p

Session 5.19
 TEACHING BASIC STATISTICS …

TWO TYPES OF ESTIMATOR

2. An INTERVAL ESTIMATOR is a formula

that gives a range of values for estimating a
parameter.
EXAMPLE:
X  d is an interval estimator of m
where d is a specified half-width of the
interval.

Session 5.20
 TEACHING BASIC STATISTICS …

POINT ESTIMATION USING SRS

X is an unbiased point estimator of m

where n

X i
X = i =1
n

and n is the sample size.

Session 5.21
 TEACHING BASIC STATISTICS …

EXAMPLE
A simple random sample of midterm exam scores of 16
STAT 1 students last term were obtained to estimate the
mean exam score of all STAT 1 students last term.
The data are as follows:
48 50 58 60 64 66 68 68
70 76 76 78 78 78 80 82
An estimate of the mean midterm score is

16
Xi 1100
X= i =1
= = 68.75
16 16

Session 5.22
 TEACHING BASIC STATISTICS …

POINT ESTIMATION USING SRSWR

s2 is an unbiased point estimator of s2 where

 ( )  i =1 i
n 2
Xi − X
n
X 2
− nX 2

s= i =1
=
n −1 n −1
Using the sample data set, earlier an
estimate of the variability of the midterm
exam scores is 10.55

Session 5.23
 TEACHING BASIC STATISTICS …

POINT ESTIMATION USING SRSWOR

s2 is an unbiased point estimator of:

N

 i
( X
i =1
− m ) 2

N −1

An unbiased estimator of s2 is N −1 2
s
N

Session 5.24
 TEACHING BASIC STATISTICS …

POINT ESTIMATION USING SRS

p̂ is an unbiased point estimator of p

where
a
pˆ =
n
and a is the number of units
possessing the characteristic of interest
out of the sample of size n.

Session 5.25
 TEACHING BASIC STATISTICS …

EXAMPLE

Using the sample data given earlier,

we determine an estimate for the
proportion of students who passed
the midterm exam with 60 as passing
score. Such estimate is given by:
a 13
pˆ = = = 0.81
n 16

Session 5.26
 TEACHING BASIC STATISTICS …

STANDARD ERROR OF THE

SAMPLE MEAN
To measure the precision of X as a
point estimator of m, we compute the
standard error of the sample mean,
denoted by s X .
The smaller the magnitude of the s X ,
the more precise is X as an estimator
of m.

Session 5.27
 TEACHING BASIC STATISTICS …

STANDARD ERROR OF THE

SAMPLE MEAN
s X is computed as
s
sX =
n if sampling is SRSWR

1N −n 
sX =s   if sampling is SRSWOR
n  N −1 

Session 5.28
 TEACHING BASIC STATISTICS …

STANDARD ERROR OF THE

SAMPLE MEAN

If s is unknown, s X is estimated by

s
sˆ X = if sampling is SRSWR
n

1N −n 
sˆ X = s 
n  N  if sampling is SRSWOR

Session 5.29
 TEACHING BASIC STATISTICS …

POINT ESTIMATOR OF m IN
STRATIFIED RANDOM SAMPLING
X st is an unbiased point estimator of m where
L
X st =  w i X i
i=1

and wi is the stratum weight computed as

Ni ; X i is the sample mean obtained from
wi =
N
each stratum, L is the number of strata, and Ni
is the number of elements in the ith stratum.
Session 5.30
 TEACHING BASIC STATISTICS …

EXAMPLE
University students were stratified according to
their affiliation in a fraternity. A researcher wishes
to estimate the grade point average of all the
students regardless of their affiliations.
Fraternity Membership wi Xi
Member 0.41 2.76
Non-member 0.59 2.54
X st =  wi X i = (0.41)(2.76) + (0.59)(2.54) = 2.63
L

i =1

Session 5.31
 TEACHING BASIC STATISTICS …

Sampling From An Infinite

Population
If (X1,X2,…,Xn) is a random sample where m=E(Xi)
and s2=Var(Xi) then an unbiased estimator of m is:
X 1 + X 2 + ... + X n
X=
n
and its precision is measured by:
s2
Var ( X ) =
n

And an unbiased estimator

n of s2 is:
(X i − X )2
S2 = i =1

n −1

Session 5.32