Anda di halaman 1dari 11

TUM School of Management

Production and Supply Chain Management


Prof Martin Grunow

Technische Universitt Mnchen

Estimating a Parameter using


Confidence Intervals
Suppose we want to estimate the mean of a normal population
using sample average X-bar as the estimator.
Using the observed value of X-bar, an interval is created such that:

P(x k x + k ) = 1
The interval (x k,x + k ) is called a (1 )100% confidence interval
for .
(1 ) is called the confidence coefficient.
The value of k is determined from the distribution of the estimator
X-bar.
2012 from "A First Course in Quality Engineering: Integrating Statistical
and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott

Quality Engineering & Management Module 5

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

C.I. for the of a Normal Population


when is known
If the population is distributed as N(,2), the sample average X-bar is
distributed as N(,2/n), where n is the sample size.
Then

( ) ~ N (0,1)

Using this result, the (1 )100% confidence interval for is obtained


as:

"x ! z !
a2
#

n, x + za/2 !

n $%

Where is the known standard deviation and z/2 is such that


P(Z > z/2) = /2
2012 from "A First Course in Quality Engineering: Integrating Statistical
and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott

Quality Engineering & Management Module 5

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

C.I. for the of a Normal Population


when is known

( ) ~ N (0,1)

"x ! z !
a2
#

n, x + za/2 !

n $%

2012 from "A First Course in Quality Engineering: Integrating Statistical


and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott

Quality Engineering & Management Module 5

10

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

C.I. for the of a Normal Population


when is known
Example:
A random sample of four bottles of fabric softener was taken from a
day's production and the amount of turbidity in them was measured
as: 12.6, 13.4, 12.8, 13.2 ppm.
Find a 99% CI for the mean turbidity of the population of bottles filled
in this line on that day.
It is known that turbidity measurements are normally distributed with a
standard deviation of = 0.3.

2012 from "A First Course in Quality Engineering: Integrating Statistical


and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott

Quality Engineering & Management Module 5

11

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

C.I. for the of a Normal Population


when is known
Need a 99% C.I.
for the mean
turbidity in (all) the
bottles of fabric
softener filled in a
line.

(1 ) = 0.99
/ 2 = 0.005
z 2 = 2.575
2012 from "A First Course in Quality Engineering: Integrating Statistical
and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott

Quality Engineering & Management Module 5

12

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

C.I. for the of a Normal Population


when is known
Example:
Need a 99% C.I. for the mean turbidity in (all) the bottles of fabric
softener filled in a line. is known to be = 0.3.
A sample of 4 bottles gave: 12.6, 13.4, 12.8, 13.2 ppm.

(1! ! ) = 0.99 " ! / 2 = 0.005 " z! 2 = 2.575


x = 52 4 = 13.0 , = 0.3
99% C.I.

"x ! z !
a2
#

n, x + za/2 !

n $%

2012 from "A First Course in Quality Engineering: Integrating Statistical


and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott

Quality Engineering & Management Module 5

13

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

C.I. for the of a Normal Population


when is known
Example:
Need a 99% C.I. for the mean turbidity in (all) the bottles of fabric
softener filled in a line. is known to be = 0.3.
A sample of 4 bottles gave: 12.6, 13.4, 12.8, 13.2 ppm.

(1! ! ) = 0.99 " ! / 2 = 0.005 " z! 2 = 2.575


x = 52 4 = 13.0 , = 0.3

99% C.I. for = 13.0 2.575 0.3

= [12.61, 13.39]
2012 from "A First Course in Quality Engineering: Integrating Statistical
and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott

4 , 13.0 + 2.575 0.3

)]

What is the interpretation


for this confidence interval?

Quality Engineering & Management Module 5

14

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

Reducing Margin of Error


Can the margin of error (and, consequently, the width) of a
confidence interval be reduced?

"x ! z !
a2
#

n, x + za/2 !

n $%

Increase Sample Size, n (but what is the problem?)


Reduce Standard Deviation, (but what is the problem?)
Decrease (1 ) : this decreases the z/2 value (but what is
the problem?)
Holly Ott

Quality Engineering & Management Module 5

15

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

Confidence Intervals
C.I. for the of a Normal
Population when is
known
C.I. for the of a Normal
Population when is not
known
C.I. for the 2 of a Normal
Population
2012 from "A First Course in Quality Engineering: Integrating Statistical
and Management Methods of Quality" by K.S. Krishnamoorthi. Reproduced
by permission of Taylor and Francis Group, LLC, a division of Informa plc.
Holly Ott

Quality Engineering & Management Module 5

16

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

Coming Up
Lecture 5.2: Hypothesis Testing

Holly Ott

Quality Engineering & Management Module 5

17

TUM School of Management


Production and Supply Chain Management
Prof Martin Grunow

Technische Universitt Mnchen

Practice
Now let's do a practice problem to calculate a
confidence interval for the mean.
Please complete the next "Practice" module
before continuing with Lecture 5.2.

Holly Ott

Quality Engineering & Management Module 5

Anda mungkin juga menyukai