Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 98

© 2003 Pearson Prentice Hall

Chapter 7

Inferences Based on a Single Sample:


Estimation with Confidence Intervals

7-1
Learning Objectives
© 2003 Pearson Prentice Hall

1. State What Is Estimated


2. Distinguish Point & Interval Estimates
3. Explain Interval Estimates
4. Compute Confidence Interval Estimates
for Population Mean & Proportion
5. Compute Sample Size

7-2
Thinking Challenge
© 2003 Pearson Prentice Hall

Suppose you’re
interested in the
average amount of
money that students
in this class (the
population) have on
them. How would
you find out?

7-3
© 2003 Pearson Prentice Hall

Introduction
to Estimation

7-4
Statistical Methods
© 2003 Pearson Prentice Hall

Statistical
Methods

Descriptive Inferential
Statistics Statistics

Hypothesis
Estimation
Testing

7-5
Estimation Process
© 2003 Pearson Prentice Hall

7-6
Estimation Process
© 2003 Pearson Prentice Hall

Population

Mean, , is
unknown

 
 


 

7-7
Estimation Process
© 2003 Pearson Prentice Hall

Population Random Sample

Mean 
Mean, , is X= 50
unknown

 
 
Sample


 

7-8
Estimation Process
© 2003 Pearson Prentice Hall

Population Random Sample


I am 95%
Mean  confident that
Mean, , is X= 50  is between
unknown
 40 & 60.
 
 
Sample


 

7-9
Unknown Population
Parameters Are Estimated
© 2003 Pearson Prentice Hall

Estimate Population with Sample


Parameter... Statistic
Mean  x
Proportion p p^
2 2
Variance  s
Differences 1 -  2 x1 -x2
7 - 10
Estimation Methods
© 2003 Pearson Prentice Hall

7 - 11
Estimation Methods
© 2003 Pearson Prentice Hall

Estimation

7 - 12
Estimation Methods
© 2003 Pearson Prentice Hall

Estimation

Point
Estimation

7 - 13
Estimation Methods
© 2003 Pearson Prentice Hall

Estimation

Point Interval
Estimation Estimation

7 - 14
© 2003 Pearson Prentice Hall

Point Estimation

7 - 15
Estimation Methods
© 2003 Pearson Prentice Hall

Estimation

Point Interval
Estimation Estimation

7 - 16
Point Estimation
© 2003 Pearson Prentice Hall

1. Provides Single Value


 Based on Observations from 1 Sample
2. Gives No Information about How Close
Value Is to the Unknown Population
Parameter
3. Example: Sample MeanX = 3 Is Point
Estimate of Unknown Population Mean

7 - 17
© 2003 Pearson Prentice Hall

Interval Estimation

7 - 18
Estimation Methods
© 2003 Pearson Prentice Hall

Estimation

Point Interval
Estimation Estimation

7 - 19
Interval Estimation
© 2003 Pearson Prentice Hall

1. Provides Range of Values


 Based on Observations from 1 Sample

2. Gives Information about Closeness to


Unknown Population Parameter
 Stated in terms of Probability

3. Example: Unknown Population Mean Lies


Between 50 & 70 with 95% Confidence

7 - 20
Key Elements of
© 2003 Pearson Prentice Hall
Interval Estimation

7 - 21
Key Elements of
© 2003 Pearson Prentice Hall
Interval Estimation

Sample statistic

(point estimate)

7 - 22
Key Elements of
© 2003 Pearson Prentice Hall
Interval Estimation

Sample statistic
Confidence
interval
(point estimate)

Confidence Confidence
limit (lower) limit (upper)
7 - 23
Key Elements of
© 2003 Pearson Prentice Hall
Interval Estimation
A probability that the population parameter
falls somewhere within the interval.

Sample statistic
Confidence
interval
(point estimate)

Confidence Confidence
limit (lower) limit (upper)
7 - 24
Confidence Limits
for Population Mean
We know the distribution of X-bar (for
large n:
 CLT says it’s normally distributed with
mean Mu)
For any z, look up Pr    X  z X 

  z X   z X
Confidence Depends on
© 2003 Pearson Prentice Hall
Interval (z)

7 - 26
Confidence Depends on
© 2003 Pearson Prentice Hall
Interval (z)

x_

X

7 - 27
Confidence Depends on
© 2003 Pearson Prentice Hall
Interval (z)
X=  ± Zx
x_

X

7 - 28
Confidence Depends on
© 2003 Pearson Prentice Hall
Interval (z)
X=  ± Zx
x_

X
-1.65x  +1.65x

90% Samples

7 - 29
Confidence Depends on
© 2003 Pearson Prentice Hall
Interval (z)
X=  ± Zx
x_

X
-1.65x  +1.65x
-1.96x +1.96x

90% Samples
95% Samples

7 - 30
Confidence Depends on
© 2003 Pearson Prentice Hall
Interval (z)
X=  ± Zx
x_

X
-2.58x -1.65x  +1.65x +2.58x
-1.96x +1.96x

90% Samples
95% Samples
99% Samples
7 - 31
Confidence Level
© 2003 Pearson Prentice Hall

1. Probability that the sample statistic


falls within interval
2. Denoted (1 - 
 Is Probability That Parameter Is Not
Within Interval
3. Typical Values Are 99%, 95%, 90%
 Is .01, .05., .10

7 - 32
Intervals
© 2003 Pearson Prentice Hall

Equivalent formulations:
  X  z X
X    z X ,   z X 
   X  z X , X  z X 
Therefore, can draw intervals around
either one, mu or x-bar:
Pr X    z X ,   z X   Pr   X  z X , X  z X 

7 - 33
But, Be Careful…
© 2003 Pearson Prentice Hall

When you write… Pr   X  z X , X  z X 


… mu and sigma are fixed.
 The experiment is to draw a sample and compute x-
bar
 Not to make a random draw of what mu is
 (Would you be selecting from all the alternate universes where
mu might be different?)
Thus, the probability is really 1 -  that if you run the
experiment again, the interval you’ll draw around
the new X-bar will contain mu
 Not the probability that mu is close to the particular x-bar that you
got from the current sample

7 - 34
Intervals &
© 2003 Pearson Prentice Hall
Confidence Level
Sampling x_
Distribution /2 /2
1 - 
of Mean _
X
x = 
Intervals (1 - ) % of
extend from intervals
X - ZX to contain .
X + ZX  % do not.
Intervals derived from
7 - 35 many samples
Factors Affecting
© 2003 Pearson Prentice Hall
Interval Width
1. Data Dispersion Intervals Extend from
 Measured by  X - ZX toX + ZX
2. Sample Size

  X =  /  n
3. Level of Confidence
(1 - )
 (Affects Z)

© 1984-1994 T/Maker Co.


7 - 36
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates

7 - 37
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

7 - 38
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

Mean

7 - 39
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

Mean Proportion

7 - 40
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

Mean Proportion Variance

7 - 41
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

Mean Proportion Variance

Known

7 - 42
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

Mean Proportion Variance

Known  Unknown

7 - 43
© 2003 Pearson Prentice Hall

Confidence Interval Estimate


Mean ( Known)

7 - 44
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

Mean Proportion Variance

Known  Unknown

7 - 45
Confidence Interval
© 2003 Pearson Prentice Hall
Mean ( Known)
1. Assumptions
 Population Standard Deviation Is Known
 Population Is Normally Distributed
 If Not Normal, Can Be Approximated by
Normal Distribution (n  30)

7 - 46
Confidence Interval
© 2003 Pearson Prentice Hall
Mean ( Known)
 1. Assumptions
 Population Standard Deviation Is Known
 Population Is Normally Distributed
 If Not Normal, Can Be Approximated by
Normal Distribution (n  30)
2. Confidence Interval Estimate
 
X  Z / 2     X  Z / 2 
n n
7 - 47
Estimation Example
© 2003 Pearson Prentice Hall
Mean ( Known)
The mean of a random sample of n = 25
isX = 50. Set up a 95% confidence
interval estimate for  if  = 10.

7 - 48
Estimation Example
© 2003 Pearson Prentice Hall
Mean ( Known)
The mean of a random sample of n = 25
isX = 50. Set up a 95% confidence
interval estimate for  if  = 10.
 
X  Z / 2     X  Z / 2 
n n
10 10
50  1.96     50  1.96 
25 25
46.08    53.92

7 - 49
Thinking Challenge
© 2003 Pearson Prentice Hall

You’re a Q/C inspector for


Gallo. The  for 2-liter
bottles is .05 liters. A
random sample of 100
bottles showedX = 1.99
liters. What is the 90%
confidence interval
estimate of the true mean
22
amount in 2-liter bottles? liter
liter
© 1984-1994 T/Maker Co.

7 - 50
Confidence Interval
© 2003 Pearson Prentice Hall
Solution*

 
X  Z / 2     X  Z / 2 
n n

.05 .05
1.99  1645
.     199
.  1.645 
100 100

1.982    1.998

7 - 51
Check Your
Understanding
You take a sample
 Compute x-bar and confidence interval of 1.96*sigma around
x-bar
1. Is Pr(mu lies in the confidence interval)=.95?
Suppose you take another sample
2. Is Pr(x-bar for new sample lies in the first confidence interval)
= .95?
3. Is Pr(confidence interval computed from the second sample
contains mu) = .95?
Suppose you combine the two samples
4. Will the confidence interval be larger or smaller than the
individual intervals computed?
5. Will the confidence level be lower or higher?
© 2003 Pearson Prentice Hall

Confidence Interval Estimate


Mean ( Unknown)

7 - 53
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

Mean Proportion Variance

Known  Unknown

7 - 54
Large Samples
© 2003 Pearson Prentice Hall

The sample variance s is a good estimator


of sigma
Carry on as before

7 - 55
Another Way To Think
© 2003 Pearson Prentice Hall
About It
Define variable X  X  X 
Z  
X  / n s/ n
X-bar is the sampling distribution of the mean of a
sample of Xs
Z is the normalized variable of X-bar
 mu= 0 and sigma = 1
By the CLT, X-bar is normally distributed (large samples)
 So Z is normally distributed
 Therefore,
       
pr  Z  Z    pr  Z   Z    pr  Z  Z    
  2   2  
7 - 56
 2  2  2
Confidence Interval
© 2003 Pearson Prentice Hall
Cookbook
X  Z  X
Interval to compute:
2
Compute sample mean, X
Z
Find 2

What is it for   .1? .05 ? .01?

Compute sample standard deviation, s


Compute  X  / n  s/ n

7 - 57
Problems for Small Samples
© 2003 Pearson Prentice Hall

X may not be normally distributed


What are the implications for Z?

s is not a good estimator of  X


n

7 - 58
Solution for Small
© 2003 Pearson Prentice Hall
Samples
1. Assumptions
 Population of X Is Normally Distributed
2. Use Student’s t Distribution
1. Define variable
X 
T
s / n-1
2. T has the Student distribution with n degrees of
freedom (When X is normally distributed)
• There’s a different Student distribution for different degrees of
freedom
• As n gets large, Student distribution approximates a normal
distribution with mean = 0 and sigma = 1

7 - 59
Student’s t Distribution
© 2003 Pearson Prentice Hall

Standard
Normal
Bell-Shaped
t (df = 13)
Symmetric
t (df = 5)
‘Fatter’ Tails

Z
t
0
7 - 60
Confidence Interval
© 2003 Pearson Prentice Hall
Mean ( Unknown)
Find t-value associated with desired
confidence level alpha
T
X  Pr T  t / 2,n1   
s/ n
1001    confidence interval is
 X t s s 
  / 2,n1 , X  t / 2,n1 
 n n

7 - 61
Student’s t Table
© 2003 Pearson Prentice Hall

7 - 62
Student’s t Table
© 2003 Pearson Prentice Hall

v t.10 t.05 t.025

1 3.078 6.314 12.706

2 1.886 2.920 4.303

3 1.638 2.353 3.182

7 - 63
Student’s t Table
© 2003 Pearson Prentice Hall

v t.10 t.05 t.025

1 3.078 6.314 12.706

2 1.886 2.920 4.303

3 1.638 2.353 3.182

t values
7 - 64
Student’s t Table
© 2003 Pearson Prentice Hall

/2
v t.10 t.05 t.025

1 3.078 6.314 12.706

2 1.886 2.920 4.303

3 1.638 2.353 3.182 /2

0 t
t values
7 - 65
Student’s t Table
© 2003 Pearson Prentice Hall

Assume:
/2 n=3
v t.10 t.05 t.025 df = n - 1 = 2
 = .10
1 3.078 6.314 12.706
/2 =.05
2 1.886 2.920 4.303

3 1.638 2.353 3.182 /2

0 t
t values
7 - 66
Student’s t Table
© 2003 Pearson Prentice Hall

Assume:
/2 n=3
v t.10 t.05 t.025 df = n - 1 = 2
 = .10
1 3.078 6.314 12.706
/2 =.05
2 1.886 2.920 4.303

3 1.638 2.353 3.182 /2

0 t
t values
7 - 67
Student’s t Table
© 2003 Pearson Prentice Hall

Assume:
/2 n=3
v t.10 t.05 t.025 df = n - 1 = 2
 = .10
1 3.078 6.314 12.706
/2 =.05
2 1.886 2.920 4.303

3 1.638 2.353 3.182 .05

0 t
t values
7 - 68
Student’s t Table
© 2003 Pearson Prentice Hall

Assume:
/2 n=3
v t.10 t.05 t.025 df = n - 1 = 2
 = .10
1 3.078 6.314 12.706
/2 =.05
2 1.886 2.920 4.303

3 1.638 2.353 3.182 .05

0 t
t values 2.920
7 - 69
Degrees of Freedom (df)
© 2003 Pearson Prentice Hall

1. Number of Observations that Are Free to


Vary After Sample Statistic Has Been
Calculated
2. Example degrees of freedom
= n -1
 Sum of 3 Numbers Is 6
= 3 -1
X1 = 1 (or Any Number)
=2
X2 = 2 (or Any Number)
X3 = 3 (Cannot Vary)
Sum = 6

7 - 70
Estimation Example
© 2003 Pearson Prentice Hall
Mean ( Unknown)
A random sample of n = 25 hasx = 50 &
s = 8. Set up a 95% confidence interval
estimate for .

7 - 71
Estimation Example
© 2003 Pearson Prentice Hall
Mean ( Unknown)
A random sample of n = 25 hasx = 50 &
s = 8. Set up a 95% confidence interval
estimate for .
S S
X  t  / 2, n 1     X  t  / 2, n 1 
n n
8 8
50  2.0639     50  2.0639 
25 25
46.69    53.30
7 - 72
Thinking Challenge
© 2003 Pearson Prentice Hall

You’re a time study


analyst in manufacturing.
You’ve recorded the
following task times (min.):
3.6, 4.2, 4.0, 3.5, 3.8, 3.1.
What is the 90%
confidence interval
estimate of the population
mean task time?

7 - 73
Confidence Interval
© 2003 Pearson Prentice Hall
Solution*
X = 3.7
S = 0.38987
n = 6, df = n - 1 = 6 - 1 = 5
S / n = 0.38987 / 6 = 1.592
t.05,5 = 2.0150
3.7 - (2.015)(1.592) 3.7 + (2.015)(1.592)
3.38  4.02
7 - 74
© 2003 Pearson Prentice Hall

Confidence Interval Estimate


of Proportion

7 - 75
Confidence Interval
© 2003 Pearson Prentice Hall
Estimates
Confidence
Intervals

Mean Proportion Variance

Known  Unknown

7 - 76
Confidence Interval
© 2003 Pearson Prentice Hall
Proportion

7 - 77
Confidence Interval
© 2003 Pearson Prentice Hall
Proportion of Population
Two Categorical Outcomes
Mean of 0-1 variables, each with prob. p
 Xi  pq
pq
count of 1' s  X i X 
X  n
n n

(Like binomial, which counted 1’s, but didn’t


divide by n)

7 - 78
Normal Approximation Can
© 2003 Pearson Prentice Hall
Be Used

 pq pq 
 X  z X , X  z X    p  z , p  z 
 n n 
 pˆ qˆ pˆ qˆ 
  pˆ  z , pˆ  z 
 n n 

7 - 79
Confidence Interval
© 2003 Pearson Prentice Hall
Proportion
1. Assumptions
 Two Categorical Outcomes
 Population Follows Binomial Distribution
 Normal Approximation Can Be Used
ˆ 1  p
p ˆ
 p3
If ˆ Does Not Include 0 or 1
n
2. Confidence Interval Estimate
  (1  p )
p   (1  p )
p
p  z 2   p  p  z 2 
n n
7 - 80
Estimation Example
© 2003 Pearson Prentice Hall
Proportion
A random sample of 400 graduates
showed 32 went to grad school. Set up a
95% confidence interval estimate for p.

7 - 81
Estimation Example
© 2003 Pearson Prentice Hall
Proportion
A random sample of 400 graduates
showed 32 went to grad school. Set up a
95% confidence interval estimate for p.
  (1  p )
p   (1  p )
p
p  Z  / 2   p  p  Z  / 2 
n n
.08  (1 .08) .08  (1 .08)
.08  1.96   p  .08  1.96 
400 400
.053  p  .107
7 - 82
Thinking Challenge
© 2003 Pearson Prentice Hall

You’re a production
manager for a newspaper.
You want to find the %
defective. Of 200
newspapers, 35 had
defects. What is the
90% confidence interval
estimate of the population
proportion defective?

7 - 83
Confidence Interval
© 2003 Pearson Prentice Hall
Solution*

  (1  p )
p   (1  p )
p
p  z / 2   p  p  z / 2 
n n

.175  (.825) .175  (.825)


.175  1.645   p  .175  1.645 
200 200

.1308  p  .2192

7 - 84
Review
© 2003 Pearson Prentice Hall

Define

Z
2

7 - 85
Review
© 2003 Pearson Prentice Hall

  Pr confidence interval will not contain  


 
 Pr  X    Z   X  (if X is normally distribute d)
 
 2 
 
or  Pr  X    T  X  (if X is normally distribute d)
 ,n 1 
 2 

7 - 86
© 2003 Pearson Prentice Hall

Finding Sample Sizes

7 - 87
Finding Sample Sizes
© 2003 Pearson Prentice Hall
for Estimating 
I don’t want to
X   Error
(1) Z  sample too much
x x or too little!


(2) Error  Z x Z
n
2 2
Z 
(3) n
Error 2
Error Is Also Called Bound, B, or
Margin of Error or Sampling Error, SE
7 - 88
Determining Sample Size
© 2003 Pearson Prentice Hall

Z is determined by desired confidence


level
But how do you determine sigma?

7 - 89
Determining Sample Size
© 2003 Pearson Prentice Hall

Z is determined by desired confidence


level
But how do you determine sigma?
 Known from previous studies
 Pilot test on a small n
 Theoretical derivation

7 - 90
Sample Size Example
© 2003 Pearson Prentice Hall

What sample size is needed to be 90%


confident of being correct within  5? A
pilot study suggested that the standard
deviation is 45.

7 - 91
Sample Size Example
© 2003 Pearson Prentice Hall

What sample size is needed to be 90%


confident of being correct within  5? A
pilot study suggested that the standard
deviation is 45.

n
Z 2

1.645 45
2 2 2
 219.2  220
Error 2
52

7 - 92
Thinking Challenge
© 2003 Pearson Prentice Hall

You work in Human


Resources at Merrill Lynch.
You plan to survey employees
to find their average medical
expenses. You want to be
95% confident that the sample
mean is within ± $50.
A pilot study showed that 
was about $400. What
sample size do you use?
7 - 93
Sample Size Solution*
© 2003 Pearson Prentice Hall

Z 2 2
n
Error 2


1.96 400
2 2

502

 245.86  246

7 - 94
Public Opinion Poll
© 2003 Pearson Prentice Hall

You want to know what percentage of the


population supports gay marriage
Within bound plus/minus 4% at 95%
confidence
How large a sample do you need?
p(1-p) achieves its maximum at p=.5
 Using a conservative bound,  p1  p   0.5

7 - 95
Public Opinion Poll
© 2003 Pearson Prentice Hall
Solution*
2 2 2 2 2
z 
1.96 pq 1.96 * .5
n  
2 2 2
B .04 .04
 600
Even if p is near .5, 95% confidence interval will be
plus/minus 4 percent
If p is much smaller, will the confidence interval be
larger or smaller? (Hint: try p=.1)
If you expect p is 10-30%, rather than 50%, do you
need a bigger or smaller sample to get 95%
confidence of plus/minus 4 percent?

7 - 96
Conclusion
© 2003 Pearson Prentice Hall

1. Stated What Is Estimated


2. Distinguished Point & Interval Estimates
3. Explained Interval Estimates
4. Computed Confidence Interval Estimates
for Population Mean & Proportion
5. Computed Sample Size

7 - 97
End of Chapter

Any blank slides that follow are


blank intentionally.

You might also like