Probability and Statistics

LECTURE 6

ESTIMATION PROBLEM

Adapted from http://www.prenhall.com/mcclave

OUTLINE

1. State What Is Estimated

2. Distinguish Point & Interval Estimates
3. Explain Interval Estimates
4. Compute Confidence Interval estimates for
Population Mean & Proportion
STATISTICAL METHODS

Statistical
Methods

Descriptive Inferential
Statistics Statistics

Hypothesis
Estimation
Testing
3
ESTIMATION PROCESS

Population Random Sample

I am 95%
Mean  confident that
Mean, , is X = 50  is between
unknown
 40 & 60.
 
 
Sample


 
4
PARAMETER VS SAMPLE STATISTIC
 Parameter is numerical descriptive measure of
population
 Sample statistic is a numerical descriptive
measure of sample
 Example

5
UNKNOWN POPULATION PARAMETERS
ARE ESTIMATED

Estimate Population with Sample

Parameter... Statistic
Mean  x
Proportion p p^
Variance 2

2
s
Differences 1 -  2 x1 -x2
ESTIMATION METHODS

Estimation

Point Interval
Estimation Estimation

7
POINT ESTIMATION
1. Provides Single Value
2. Example:
3. Disadvantage: Gives No Information about
How Close Value Is to the Unknown
Population Parameter

8
INTERVAL ESTIMATION

1. Provides Range of Values

2. Gives Information about Closeness to Unknown
Population Parameter
 Stated in terms of Probability
 Knowing Exact Closeness Requires Knowing Unknown
Population Parameter

3. Example: Unknown Population Mean is contained in

the interval (50;70) with 95% Confidence

9
KEY ELEMENTS OF INTERVAL ESTIMATION

A probability that the interval contains the

population parameter.
Sample statistic
(point estimate)
Confidence interval

Confidence Confidence
limit (lower) limit (upper) 10
CONFIDENCE INTERVAL (CI) ESTIMATES

Confidence
Intervals

Mean Proportion

Known  Unknown

11
CONFIDENCE INTERVAL FOR
POPULATION MEAN ( KNOWN)
1. Assumptions
 Population Standard Deviation Is Known
 Population Is Normally Distributed
 If Population Is Not Normal, Large Sample Size (so
that the CLT holds) Is Required (In This Case The
Sampling Distribution Of 𝑋ത Is Approximately
Normal)
2. Confidence Interval Estimate
𝜎 𝜎
ത
𝑋 − 𝑧𝛼 ത
≤ 𝜇 ≤ 𝑋 + 𝑧𝛼
2 𝑛 2 𝑛
𝛼
where 𝑧𝛼 satisfies that: 𝑃 𝑧 > 𝑧𝛼 =
2 2 2 12
PROOF OF THE CONFIDENCE INTERVAL
(CI) FORMULA
 This proof is optional
 Start with sampling distribution of sample
mean. Recall that under some conditions, this
distribution is normal or approx. normal
 Convert to Z distribution

13
PROOF OF THE CONFIDENCE INTERVAL
(CI) FORMULA

α/2
1-α
x

α/2
1-α
Z 14
PROOF OF THE CONFIDENCE INTERVAL
(CI) FORMULA

 Let’s define a new notation:

 Zα/2 is the Z value such that the area to its right is equal
to α/2
 So we have figure:

Z 15

-Zα/2 0 Zα/2
PROOF OF THE CONFIDENCE INTERVAL
(CI) FORMULA
P(-Zα/2 < Z < Zα/2) = ?

Replace

Rearranging the expression, we should obtain an

probability equation for the CI.

16
CONFIDENCE LEVEL

1. Probability that the Interval contains

Unknown Population Parameter
2. Denoted 1 - 
 Is Probability That Interval does NOT contain
Parameter
3. Typical Values Are 99%, 95%, 90%

17
Z AND 

/2 /2
1-


Z / 2 0 Z / 2
As  is the probability that the interval does not
contain , we construct an interval that places area
/2 in each tail. 𝑧𝛼 is the z value such that the area
2
/2 lies to its right.
18
Z AND 

/2 /2
1-


Z / 2 0 Z / 2
If we know , we can find 𝑧𝛼 using Table 1 in
2
Appendix of your textbook.
Example:  = 0.1 → /2 = 0.05 → P(Z < 𝑧0.05 ) = 0.95
Look up table 1, and we find 𝑧0.05 = 1.645 19
HOW TO LOOK UP TABLE 1

20
 Z AND 

 Note that we actually find 2 values for Z by looking

up Table 1:
 With P = 0.9495, we have Z = 1.64

 With P = 0.9505, we have Z = 1.65.

 Taking the average of the two values, we have

𝑧0.05 = 1.645.

21
FACTORS AFFECTING INTERVAL WIDTH

1. Data variability
measured by  Intervals Extend from
𝜎 X - ZX toX + ZX
2. Sample Size 𝜎𝑋ത =
𝑛
3. Level of Confidence
(1 - ) Affects Z

22

© 1984-1994 T/Maker Co.

INTERVALS & CONFIDENCE LEVEL

x_
Sampling /2 /2
Distribution 1 -
_
of Mean X
x = 
Intervals 100(1 - ) %
extend from of intervals
X - ZX to contain .
X + ZX 100 % do
not.
Large number of intervals 23
ESTIMATION EXAMPLE MEAN ( KNOWN)

The mean of a random sample of n = 25 isX = 50. Set

up a 95% confidence interval estimate for  if  = 10.
Assume the population is normally distributed.

24
ESTIMATION EXAMPLE SOLUTION

 
X  Z / 2     X  Z / 2 
n n
10 10
50  1.96     50  1.96 
25 25
46.08    53.92

25
CONFIDENCE INTERVAL FOR
POPULATION MEAN ( UNKNOWN)

1. Assumptions
 Population Standard Deviation Is Unknown
 Population is Normally Distributed

2. Use Student’s t Distribution

3. Confidence Interval Estimate
𝑠 𝑠
𝑋ത − 𝑡𝛼𝑛−1, . ≤ 𝜇 ≤ 𝑋ത + 𝑡𝛼𝑛−1, .
2 𝑛 2 𝑛

26
T STATISTIC
𝑋ത − 𝜇
𝑡= 𝑠
𝑛

 If we are sampling from a normal distribution, the t

statistic has a sampling distribution very much like
that of the z statistic: mound-shaped, symmetric,
with mean 0.

 The main difference between the sampling

distributions of t and z is that the t statistic is more
variable than the z.
27
DEGREES OF FREEDOM (DF)

1. Number of Values that Are Free to Vary when

Calculating a Sample Statistic.
2. Degrees of freedom of the t statistic (for one-
sample case) is n1.
3. Example

28
DEGREES OF FREEDOM (DF)

The smaller the number of degrees of freedom

associated with the t statistic, the more spread out
the sampling distribution of t will be.

29
STUDENT’S T DISTRIBUTION

Standard
Normal

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

Z
t
0 30
STUDENT’S T TABLE (TABLE 2 IN
APPENDIX OF TEXTBOOK)
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
STUDENT’S T TABLE
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
ESTIMATION EXAMPLE MEAN
( UNKNOWN)
A random sample of n = 25 hasx = 50 & s = 8.
Set up a 95% confidence interval estimate for .
Assume normal population.

33
ESTIMATION EXAMPLE SOLUTION

S S
X  t / 2,n 1     X  t / 2,n 1 
n n
8 8
50  2.064     50  2.064 
25 25
46.69    53.30

34
CONFIDENCE INTERVAL ESTIMATES

Confidence
Intervals

Mean Proportion

Known  Unknown

35
CONFIDENCE INTERVAL PROPORTION
1. Assumptions
o Two Categorical Outcomes
o Population Follows Binomial Distribution
o Normal Approximation To The Sampling Distribution Of 𝑝
ො
Can Be Used
𝑛 𝑝Ƹ ≥ 5 and 𝑛 1 − 𝑝Ƹ ≥ 5

2. Confidence Interval Estimate

𝑝Ƹ 1 − 𝑝Ƹ 𝑝Ƹ 1 − 𝑝Ƹ
𝑝Ƹ − 𝑧𝛼 ≤ 𝑝 ≤ 𝑝Ƹ + 𝑧𝛼
2 𝑛 2 𝑛

36
ESTIMATION EXAMPLE PROPORTION
A random sample of 400 graduates showed 32
went to grad school. Set up a 95% confidence
interval estimate for p.

37
ESTIMATION EXAMPLE SOLUTION

  (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

38
PRECISION OF ESTIMATE

 The smaller the width of CI, the higher the

precision of estimation.
 What are the ways to make the estimate more
precise?
 In practice, if we choose to increase sample size to
increase precision, we should also take into account
the costs of study.

39
CI FOR FINITE POPULATIONS
 In cases where n/N > 0.05, the formulas for CI
must be adjusted (as the standard error of
estimate changes)
 Please consult your textbook or reference books if
you are interested in the adjusted formulas.
Knowing these formulas is not required for the
test/exam.

40
CONCLUSION

1. State What Is Estimated

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

41