You are on page 1of 30

BMS1024 MANAGERIAL STATISTICS

BMS1024
MANAGERIAL
STATISTICS
Estimation and Sample Size
Determination
BMS1024 MANAGERIAL STATISTICS
Topic Outline
1. Confidence Intervals for the Population Mean,
when Population Standard Deviation is
Known
when Population Standard Deviation is
Unknown
2. Confidence Intervals for the Population Proportion,
p
3. Determining the Required Sample Size
BMS1024 MANAGERIAL STATISTICS
Point Estimates
A point estimate is a single number.
For the population mean (and population standard deviation), a point
estimate is the sample mean (and sample standard deviation).
A confidence interval provides additional information about
variability
Point Estimate
Lower Confidence
Limit (LCL)
Upper Confidence
Limit (UCL)
Width of
confidence interval
BMS1024 MANAGERIAL STATISTICS
Confidence Interval Estimates
A confidence interval gives a range estimate of
values:
Gives information about closeness to unknown
population parameters
Stated in terms of level of confidence
Ex. 90% confidence, 95% confidence, 99% confidence
Can never be 100% confident
BMS1024 MANAGERIAL STATISTICS
Confidence Level
Suppose confidence level = 95%
Also written (1 - o) = .95
A relative frequency interpretation:
In the long run, 95% of all the confidence
intervals that can be constructed will contain
the unknown true parameter
A specific interval either will contain or will
not contain the true parameter
BMS1024 MANAGERIAL STATISTICS
Confidence Interval for
( Known)
Assumptions:
Population standard deviation is known
Population is normally distributed

Confidence interval estimate:




(where Z is the standardized normal distribution critical value
for a probability of /2 in each tail)
|
.
|

\
|
=
n

Z X
2 / o
BMS1024 MANAGERIAL STATISTICS
Finding the Critical Value, Z
Consider a 95% confidence interval = 95% CI:
Z= -1.96 Z= 1.96
.95 1 = o
025
2
.

= 025
2
.

=
Lower
Confidence
Limit (LCL)
Upper
Confidence
Limit (UCL)
Z units:
X units:
Point Estimate
0
BMS1024 MANAGERIAL STATISTICS
Confidence Interval for
( Known)
To find the lower confidence limit (LCL):



To find the upper confidence limit (UCL):



(where Z is the standardized normal distribution critical
value for a probability of /2 in each tail)
|
.
|

\
|
=
n

Z X
2 / o
|
.
|

\
|
+ =
n

Z X
2 / o
BMS1024 MANAGERIAL STATISTICS
Finding the Critical Value, Z
Commonly used confidence levels are 90%, 95%,
and 99%
Confidence
Level
Confidence
Coefficient

Z value
1.28
1.645
1.96
2.33
2.58
3.08
3.27
.80
.90
.95
.98
.99
.998
.999
80%
90%
95%
98%
99%
99.8%
99.9%
BMS1024 MANAGERIAL STATISTICS
Confidence Interval for
( Known) Example
A sample of 11 circuits from a large normal
population has a mean resistance of 2.20 ohms.
We know from past testing that the population
standard deviation is .35 ohms.

Determine a 95% confidence interval for the true
mean resistance of the population.
BMS1024 MANAGERIAL STATISTICS
Confidence Interval for
( Known) Example
2.4068) , (1.9932
.2068 2.20
) 11 (.35/ 1.96 2.20
n


2 / 05 . 0
=
=
= Z X
Interpretation: We are 95% confident that the true mean
resistance is between 1.9932 and 2.4068 ohms.
Although the true mean may or may not be in this interval, 95%
of intervals formed in this manner will contain the true mean.
BMS1024 MANAGERIAL STATISTICS
Confidence Interval for
( Unknown)
If the population standard deviation is
unknown, we can substitute the sample
standard deviation, s
This introduces extra uncertainty, since s is
variable from sample to sample
So we use the t-distribution instead of the
normal distribution
BMS1024 MANAGERIAL STATISTICS
Confidence Interval for
( Unknown)
Assumptions
Population standard deviation is unknown
Population is normally distributed

Use t-Distribution
Confidence Interval Estimate:


(where t is the critical value of the t distribution with n-1
d.f. and an area of /2 in each tail)
|
.
|

\
|
=
n
s
t X
n-1 , 2 / o

BMS1024 MANAGERIAL STATISTICS


Confidence Interval for
( Unknown)
To find the lower confidence limit (LCL):



To find the upper confidence limit (UCL):



(where t is t-critical value for a probability of /2 in
each tail and d.f. = n - 1)
|
.
|

\
|
=

n
s
t X
n 1 , 2 / o
|
.
|

\
|
+ =

n
s
t X
n 1 , 2 / o
BMS1024 MANAGERIAL STATISTICS
Properties of t-Distribution
t
0
t (df = 5)
t (df = 13)
t-distributions are bell-shaped
and symmetric, but have
fatter tails than the normal
Standard
Normal
(t with df = )
Note: t Z as n increases
BMS1024 MANAGERIAL STATISTICS
t-table: Table 10 from Cambridge
Statistical Tables
d.f. = n - 1
Let: n = 3
df = n - 1 = 2
90% CI ~ o = .10
o/2 =.05
o/2 = .05
t
0
2.920
t-values based on d.f. & o/2 (%)
o/2 (expressed in %)
BMS1024 MANAGERIAL STATISTICS
Confidence Interval for
( Unknown) Example
A random sample of n = 25 has X = 50 and s = 8. Form a 95%
confidence interval for
d.f. = n 1 = 24, so the confidence interval is
|
.
|

\
|
=
|
.
|

\
|
=
|
.
|

\
|
=
25
8
0639 2 50
25
8
50
24 , 2 / 05 . 0
1 2
) . (
t
n
s
t X
, n- /

= (46.698 , 53.302)
Interpretation: We are 95%
confident that the true mean is
between 46.369 and 53.302.
BMS1024 MANAGERIAL STATISTICS
Confidence Intervals for the
Population Proportion, p
An interval estimate for
the population
proportion (p) can be
calculated by:
n
) p ( p
Z p p
. .
.

=
1
2 / o
where
Z is the standardized normal value for the level of confidence
desired
p-hat is the sample proportion
n is the sample size
Upper and lower confidence limits for
the population proportion are
calculated with the formula:
BMS1024 MANAGERIAL STATISTICS
Confidence Intervals for the
Population Proportion, Example
A random sample of 100 people shows that 25 have opened a
current account this year. Form a 95% confidence interval for
the true proportion of the population who have opened a current
account.
( )
) (. .
. /
n
) p ( p
Z p p
100
75 25
96 1 100 25
1
2 /
=

=
. .
.
o
) . , . (
) (. . .
3349 0 1651 0
0433 96 1 25
=
=
Interpretation: We are 95%
confident that the true proportion
of population who have opened a
current account is between
16.51% and 33.49%.
BMS1024 MANAGERIAL STATISTICS
Summary of Confidence Interval
The width of the confidence interval is wider
when:
The confidence level increases
Sample size (n) decreases
Standard deviation increases

The width of the confidence interval is
narrower when?
BMS1024 MANAGERIAL STATISTICS
Topic Recap
|
.
|

\
|
=
n

Z X
2 / o
|
.
|

\
|
=
n
s
t X
n-1 , 2 / o

Confidence Interval
Mean Proportion
known unknown
n
) p ( p
Z p p
. .
.

=
1
2 / o
BMS1024 MANAGERIAL STATISTICS
Determining Sample Size (Mean)
To determine the
appropriate sample size,
these information is
required:
Confidence level (CI)

E = Margin of Error or also
known as Error in
Estimation or Bound of
Error
|
.
|

\
|
=
n

Z X
2 / o
n

Z E =
Now solve for n :
2
2
2 2


> >
E
Z
E
Z
n
o
Always round up n to the next integer!
BMS1024 MANAGERIAL STATISTICS
Determining Sample Size (Mean)
Example
If o = 45, what sample size is needed to estimate
the mean within 5 with 90% confidence?
19 219
5
45 645 1
2
2 2
2
2 2
.
) ( ) . (
E
Z
n = = >
So the required sample size is n at least 220
BMS1024 MANAGERIAL STATISTICS
Determining Sample Size (Proportion)
To determine the appropriate
sample size, these information
is required:
Confidence level (CI)

p^ = p-hat = sample proportion

E = Margin of Error or also
known as Error in Estimation or
Bound of Error
n
p p
Z p p
) 1 (
2 /
. .
.

=
o
n
p p
Z E
) 1 (
. .

=
Now solve for n :
2
2
) 1 (
E
p p Z
n
. .

>
Always round up n to the next integer!
BMS1024 MANAGERIAL STATISTICS
Determining Sample Size
(Proportion)
Sample proportion in the success category is denoted by p^




If the value of p^ is unknown, use p^ = 0.5. It will
give the maximum number of required sample
size.
size sample
sample in successes of number
= =
n
X
p
^
BMS1024 MANAGERIAL STATISTICS
Determining Sample Size (Proportion)
Example
How large a sample would be necessary to
estimate the true proportion defective in a
large population within 3%, with 95%
confidence?
(Assume a pilot sample yields ) 12 . 0 =
.
p
BMS1024 MANAGERIAL STATISTICS
Determining Sample Size (Proportion)
Example
Solution:
For 95% confidence, use Z = 1.96
E = .03
= sample proportion
So use n > 451
450.75
(.03)
.12) (.12)(1 (1.96) ) 1 (
2
2
2
2
=

>
. .
E
p p Z
n
12 . 0 =
.
p
BMS1024 MANAGERIAL STATISTICS
Working in Class: Example
Many companies have been experiencing downsizing.
There is some feeling that when rumors of a downsize
begins to float through an organization, employees
begin to use their sick days at a faster rate than normal
for fear of losing them. Fifty employees were surveyed
and asked if they would use their sick days freely in the
face a potential downsizing. Of the 50 employees, 20
said they would.
BMS1024 MANAGERIAL STATISTICS
Working in Class: Example
a) Find a 95% confidence interval for the proportion of
employees who would begin to use their sick days freely when
faced with a potential downsizing.

b) What sample size should be taken if the error in estimation is
to be not more than 6%?

c) What sample size would be taken in the absence of any
information on the proportion of employees who would freely
use their sick days when faced with a potential downsizing?
BMS1024 MANAGERIAL STATISTICS
At the end of the lesson, you should be
able to:
To construct and interpret confidence interval
estimates for the mean and the proportion
How to determine the sample size necessary to
develop a confidence interval for the mean or
proportion

You might also like