You are on page 1of 50

www.covenantuniversity.edu.

ng

Raising a new Generation of Leaders

PROBABILITY AND
STATISTICS
MODULE THREE

By: Dr. O. A. ODETUNMIBI


ESTIMATION

2
INTRODUCTION
Estimation is a process whereby we select a
random sample from a population and use a
sample statistic to estimate a population
parameter.
3
A measure calculated from population data is
called Parameter.
A measure calculated from sample data is called
Statistic.
Population parameters Sample statistic
Size N n
Mean  y
Variance 2 s 2

Standard deviation  S

4
Estimation is subdivided into two, which are: Point
estimation and interval estimation.

In point estimation, a single parameter is computed


from the sample information and is used as an
estimate of the value.

In interval estimation, two points or a critical point


are used defining an interval on the real number line
which hopefully will contain the value.
5
POINT ESTIMATION
A point estimate is a single value and has the
advantage of being very precise but there is no
information about its reliability

This single number is arrived at by selecting a


specific function of the sample values called the
estimator of θ. Once the form of the function is
selected, an estimate of θ is achieved by evaluating
the function at the sample values: x1, x2, . . ., xn.
6
PROPERTIES OF AN ESTIMATOR:
a). Unbiasedness
b). Efficiency
c). Consistency
d). Sufficiency
7
UNBIASEDNESS (UNBIASED ESTIMATOR)
An estimator is said to be unbiased if the expected
value of an estimator equals the population
parameter. That is, the average values of the
estimators equal the actual parameter. For example,
show that x is an unbiased estimator of 

8
EXAMPLE I
If X1, X2 and X3 are independent random variables
having a Poisson distribution with parameter λ. Show
that the following estimators of λ are unbiased.

1. X1  X 2  X 3 2. X 1  2 X 2  3 X 3 3. X 1  2 X 2  X 3
3 6 4

9
EXAMPLE II
Assume that X1 to X4 are independent random
observations from a Normal distribution having
mean µ and variance σ2. Determine which of the
following estimators of µ are unbiased.
1. ˆ 1 1 2. ˆ  X1  X 2  X 3  X 4
1   X1  X 2    2 X 3  X 4  2
4 6 2 3 4 5
10
EXERCISE
If X1, X2 , X3 and X4 are independent random
variables having the Bernoulli distribution with
parameter θ. Show that the following estimators of θ
are unbiased.
1. X1  X 2  X 3  X 4 2. 3 X1  2 X 2  3 X 3
4 8
11
EFFICIENCY:
The property of efficiency implies that the variance of
the estimators should be small. Having a small
variance does not make an estimator a good one until
the estimator is unbiased.
The property of efficiency of an estimator is defined by
comparing its variance to the variance of all other
estimators. The most efficient estimator among a
group of unbiased estimators is the one with the
smallest variance.

12
CONSISTENCY
It is also desirable that an estimator tend to lie nearer to
the population characteristics as the sample size
becomes larger.
An estimator is said to be consistent if the larger the
sample size, the more likely it is that the estimator will be
closed to θ (the population parameter/ estimate).

13
SUFFICIENCY
A sufficiency statistic is such that the sample mean is an
estimator that utilizes all the information about the
population parameter that the sample can provide. That
is, the sufficient estimator takes into account, each of the
sample characteristics as well as all the information that
is provided by these observations.
14
METHODS OF ESTIMATING PARAMETERS

1) Method of Moments

2) Method of Maximum Likelihood

15
METHOD OF MOMENTS
In the method of moments, if there are k
parameters that have to be estimated, the first k
sample moments are set to equal to the first k
population moments, that are given in terms of the
unknown parameters. These k – equations are then
solved simultaneously for the unknown parameters.
16
E ( X )   x f ( x)
k For a discrete
distribution

E ( X )   x f ( x)dx
k For a continuous
distribution

17
METHOD OF MAXIMUM LIKELIHOOD
This method is used for finding estimators that
satisfies some (but not necessarily all) of the
criteria discussed previously i.e. properties of
estimators.
It is used to find point estimates of parameters.
The method of maximum likelihood obtains as a
parameter estimate, that value for which the
probability of obtaining the set of sample values is
a maximum.
18
Advantages of Maximum Likelihood Estimation
(a). They are asymptotically efficient.
(b). If a sufficient estimator for a parameter exist,
the maximum likelihood estimator exist.
(c). MLE are not necessarily biased.

19
19
INTERVAL ESTIMATION
Interval estimation also known as confidence
interval estimate is an estimation of population
parameter giving by two numerical values,
between which the parameter may be considered
to lie with a given probability.

20
The purpose of interval estimation is to construct
range of values within which the population
parameters are expected to lie with a given
probability based on the result of a random sample.
We wish to estimate a parameter say µ by an interval
a < µ < b where a and b are obtained from the sample
observations.
21
A. Confidence Interval for µ
when σ is known and n > 30
If x is the mean of the random sample n
from a population with known variance
σ2, then (1 – α)100% confidence interval
for µ is given by: x  Z  
2 n
22
 
x  Z    x  Z
2 n 2 n

23
EXAMPLE III
In a study, it was observed that in a sample of 50
microdrills drilling a low-carbon alloy steel, the
average lifetime (expressed as the number of holes
drilled before failure) was 12.68 with a standard
deviation of 6.83. Find a 95% confidence interval for the
mean lifetime of microdrills under these conditions.
24
EXAMPLE IV
In a sample of 100 steel wires, the average breaking
strength is 50KN with a standard deviation of 2KN.
Find a 95% confidence interval for the mean
breaking strength of the type of wire.

25
B. Confidence Interval for µ
when σ is unknown and n < 30
If x is the mean of the random sample n from
a population with unknown variance σ , then
2

(1 – α)100% confidence interval for µ is


given by: x  t s

, n 1 n
2

26
EXAMPLE V
The contents of 7 similar containers of a
commercial soap are 9.8, 10.4, 10.2, 9.8, 10.2, 10.0
and 9.6 litres. Find the 95% confidence interval for
the mean soap content of all such containers
assuming an approximate normal distribution for
container contents.
27
EXAMPLE VI
A sample of 12 measurements of the breaking
strength of cotton threads gave a mean of 7.38cm
and a standard deviation of 1.24cm. Find the 95%
confidence limits for the actual mean breaking
strength.

28
C. Confidence Interval for 1  2
when 1 and  2 are known and n > 30
2 2

 2
 2
x1  x2  Z  1
 2

2 n1 n2

 
2 2
  2 2
x1  x2  Z  1
  1  2  x1  x2  Z 
2
1 2

2 n1 n2 2 n1 n2

29
EXAMPLE VII
A standardized mathematics test was given to 75 boys
and 50 girls, the boys made an average grade of 82
and a standard deviation of 8 while the girls made an
average grade of 76 with a standard deviation of 6.
Find the 96% confidence interval of the difference
between the mean of the boys and the girls.
30
EXAMPLE VIII
The data below was as a result of quality assurance
conducted by an engineer to test the efficiency of
two brands of capacitors. The capacitance (in
ampere –hours) are measured and recorded in the
table below.

31
Brand Number of Average Standard
Samples Deviation
A 48 178 14
B 52 170 15

Find the 95% confidence interval for the difference


of the two means that is µ1 - µ2.

32
D. Confidence Interval for 1  2
when  and  are unknown & n 30
2
1
2
2

(n1  1) s  (n1  1) s  1
2 2
1 
x1  x2  t 1 2
  
2
, n1  n2  2  n1  n2  2  n1 n2 

33
EXAMPLE IX
In a bid to award scholarships to bright pupils, test
on intelligence quotient (I.Q) was conducted in two
local government areas (LGA). For LGA-1, 18
students showed a mean I.Q values of 110 with a
standard deviation of 12, while LGA-2, 15 students
displayed I.Q values of 118 with a standard
deviation of 10. Construct a 90% confidence
interval for the question.
34
E. Confidence Interval for Proportion

pˆ (1  pˆ ) x
pˆ  Z  where pˆ 
2 n n

35
EXAMPLE X
In a simple random sample of 100 automobiles
registered in a certain state, 48 of them were found to
have emissions levels that exceed the state standard.
1). Find the 95% confidence interval for the
proportion of automobiles in the state whose
emissions exceed the stipulated state standards.
36
2). Find the 95% confidence interval for the
proportion of automobiles in the state whose
emissions does not exceed the stipulated state
standards.

37
EXAMPLE XI
Stainless steels can be susceptible to stress
corrosion cracking under certain conditions. A
materials engineer is interested in determining the
proportion of steel alloy failures that are due to
stress corrosion cracking. In a sample of 200
failures, 30 of them were caused by steel corrosion
cracking. Find the 95% confidence interval for the
proportion of failures caused by stress corrosion
cracking.
38
F. Confidence Interval for D  1  2
for paired obsevations when n< 30
The methods discussed so far for finding
confidence intervals on the basis of two
samples required that the samples be
independent. In some cases, it is better to
design an experiment so that each item in one
sample is paired with an item in the other.
39
sd sd
d  t   D  d  t
, n 1 n , n 1 n
2 2

d i n d    d 
2 2

where d  i 1
and sd 
n n  n  1

40
EXAMPLE XII
A taxi Company is trying to decide whether to
purchase brand A or brand B tyres for its fleet of
taxis. The stress test was conducted using 10 of
each brand and the results are given in the
following table.

41
Brand A 76 60 85 58 91 75 82 64 79 88
Brand B 81 52 87 70 86 77 90 63 85 83

Find a 99% confidence intervals for the true


differences in the tyres.

42
EXAMPLE XIII
For a sample of nine automobiles, the mileage (in
1000s of miles) at which the original front brake pads
were worn to 10% of their original thickness was
measured, as was the mileage at which the original
rear brake pads were worn to 10% of their original
thickness. The results are given in the table below.
43
Automobile 1 2 3 4 5 6 7 8 9
Front 32.8 26.6 35.6 36.4 29.2 40.9 40.9 34.8 36.6
Rear 41.2 35.2 46.1 46.0 39.9 51.7 51.6 46.1 47.3

Find the 95% confidence interval for the difference


in mean lifetime between the front and rear brake
pads.

44
EXERCISE 1
A thread manufacturer tests a sample of eight lengths of
a certain type of thread made of blended materials
and obtains a mean tensile strength of 8.2 lb with
standard deviation 0.06 lb. Assuming tensile strengths
are normally distributed, construct a 90% confidence
interval for the mean tensile strength of this thread.

45
EXERCISE 2
Records of 40 used passenger cars and 40 used pickup
trucks (none used commercially) were randomly
selected to investigate whether there was any
difference in the mean time in years that they were kept
by the original owner before being sold. For cars the
mean was 5.3 years with standard deviation 2.2 years.
For pickup trucks the mean was 7.1 years with standard
deviation 3.0 years. Construct 95% confidence interval
for the difference in the means based on these data.
46
EXERCISE 3
A university administrator wishes to know if there
is a difference in average starting salary for
graduates with master’s degrees in engineering
and those with master’s degrees in business.
Fifteen recent graduates with master’s degree in
engineering and 11 with master’s degrees in
business were surveyed and the results are
summarized in the table below.
47
n Mean Standard
Deviation
Engineering 15 68535 1627
Business 11 63230 2033

Construct 99% confidence interval for the


difference in the population means based on these
data.
48
EXERCISE 4
A tire manufacturer is interested in testing the fuel
economy for two different tread patterns. Tires of
each tread type are driven for 1000 miles on each of
10 different cars. The mileages in mpg are presented
in the following table.
Car 1 2 3 4 5 6 7 8 9 10
Tread A 24.1 24.5 22.3 26.1 22.6 23.3 22.4 19.9 21.5 24.6
Tread B 20.3 22.5 19.7 23.2 20.4 23.5 21.9 18.6 22.4 24.9

Find a 99% confidence interval for the mean difference


in fuel economy.
49
EXERCISE 5
A soft drink manufacturer purchases aluminum cans
from an outside vendor . A random sample of 90
cans is selected from a large shipment, and each is
tested for strength by applying an increasing load to
the side of the can until it punctures. Out of the 90
cans, 62 meet the specifications for puncture
resistance. Find the 95% confidence interval for the
proportion of cans in the shipment that meet the
specifications.
50

You might also like