You are on page 1of 6

1

Chapter 8: Estimation of Mean & Proportion
 Statistics is the science of learning from data
 Collecting data
 Summarizing and analyzing data
 Making inference & drawing conclusion, Interpretation
 Theoretical foundation – probability theory
 Statistics has two big topics in this class
 Summary statistics & Inferential statistics
 Inferential statistics
 Estimation of population parameters, including point and interval
estimate
 Hypothesis testing of population parameters
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-1
8.1 Estimation: An Introduction
 What is estimation of a parameter?
 The assignment of value(s) to the parameter based on a value of the
corresponding sample statistic is called estimation.
 Estimator – unobserved sample statistic, is a random variable
 Estimate – observed sample statistic, a number.
 The estimation procedure involves the following steps
 Select a sample.
 Collect the required information from the members of the sample.
 Calculate the value of the sample statistic.
 Assign value(s) to the corresponding population parameter
 Examples
 Average life time of certain type of light bulbs?
 Random sample: 20 light bulbs
 What is estimator?
 What is estimate?
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-2
8.2 Point & Interval Estimates
 Point estimate
 The value of a sample statistic that is used to estimate a parameter is
called a point estimate
 Why it is called “point” estimate?
 How to obtain the point estimate?
 Interval estimate
 Instead of one single value, an interval (two ends), which is
constructed around the point estimate, is used
 It is also stated that how likely (95%, say) the interval contains the
corresponding parameter
 How to construct such an interval estimate? 2 values 1 probability
 Why do we need two estimates?
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-3
Example of Estimation
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-4
Instead of a point estimate, we provide interval estimate,
i.e., point estimate ±a number.
How to find the “a number”?
2
8.3 Estimation of a Population Mean: o Known
 Three possible cases with known σ
 Case I: the population from which the sample is selected is normally
distributed. In this case, sample size is not matter.
 Case II: the population is not normal, but the sample size is large (i.e.,
n ≥ 30)
 Case III: the population is not normal AND the sample size is small
(i.e., n < 30)
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-5
Steps to Construct Confidence Interval on μ
 Fundamental idea
 Two end points and a probability (confidence)
 Probability comes from a known distribution
 Sample mean’s distribution
 Steps to obtain confidence interval on μ
 Point estimate -----------------
 Critical value ------------------- z from standard normal distribution
 Margin of error
 Confidence interval ----------- ± E
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-6
2
~ ( , ) ~ (0,1)
/
Find such that ( ) 0.95 (or in general 1 )
Solve for from
/
X
X N N
n n
z P z Z z
X
z z
n
µ o
µ
o
o
µ
o
µ
÷
¬
÷ s s = ÷
÷
÷ s s
X
E z
n
o
= ×
X
Finding z for a 100(1 – α)% Confidence Level
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-7
Left tail of z = 1 – α/2
Table, Excel (normSinv),
or calculator (invNorm)
How to find z?
Example 8-1– Normal Population
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-8
A publishing company has just published a new college textbook. Before the
company decides the price at which to sell this textbook, it wants to know the
average price of all such textbooks in the market. The research department at
the company took a sample of 25 comparable textbooks and collected
information on their prices. This information produced a mean price of \$145
for this sample. It is known that the standard deviation of the prices of all such
textbooks is \$35 and the population of such prices is normal.
a). What is the point estimate of the mean price of all such college textbooks?
b). Construct a 90% confidence interval for the mean price of all such college
textbooks.
Solution: a). Point estimate of mean µ = sample mean = \$145
b). 90% C.I. for mean µ:
Point estimate = \$145
90% => z, i.e., P(Z < z) = 0.95. z = 1.65
E = z × = 1.65×35/5 = 1.65×7 = 11.55
90% C.I. for µ is 145 ± 11.55
x
x
3
Example 8.2 – Large Sample
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-9
In a 2009 survey by I-Pension LLC, adults with annual household
incomes of \$50,000 to \$125,000 were asked about the average
time they spend reviewing their 401(k) statements. Of the adults
surveyed, about 72% said that they spend less than 5 minutes, and
27% said that they spend 5 to 10 minutes to review their 401(k)
statements. Suppose a random sample of 400 adults of all income
levels, who have 401(k) statements, were asked. The sample
produced a mean of 8 minutes. Assume that the standard
deviation of such times for all 401(k) account holders is 2.20
minutes. Construct a 99% confidence interval for the mean time
spent by all 401(k) holders reviewing their 401(k) statements.
Solution: 99% C.I. for mean µ:
1. Point estimate: 8
2. 99% => z, i.e., P(Z < z) = 0.995. z = 2.58
3. E = z × = 2.58×2.2/20 = 2.58 × 0.11 = 0.284
4. 99% C.I. for µ is 8 ± 0.284
Sample Size for the Estimation of Mean
 Width of a Confidence Interval
 The width of a confidence interval depends on the size of the margin
of error, . Hence, the width of a confidence interval
can be controlled using
 The value of z, which depends on the confidence interval
 The sample size, n
 Determining the Sample Size for the Estimation of μ
 Given the confidence level and the standard deviation of the
population, the sample size that will produce a predetermined margin
of error E of the confidence interval estimate of μ is
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-10
E z
n
o
= ×
o
=
2 2
2
E
z
n
Example 8-3
 An alumni association wants to estimate the mean debt of
this year’s college graduates. It is known that the population
standard deviation of the debts of this year’s college
graduates is \$11,800. How large a sample should be
selected so that the estimate with a 99% confidence level is
within \$800 of the population mean?
 Solution
 99% => z = 2.58
 E = 800
 Standard deviation σ = 11800
Therefore, n = z
2
σ
2
/ E
2
= 1448.18 ≈1449
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-11
8.4 Estimation of a Population Mean: σ unknown
 Three possible cases with unknown σ
 Case I: the population from which the sample is selected is normally
distributed. In this case, sample size is not matter.
 Case II: the population is not normal, but the sample size is large
(i.e., n ≥ 30)
 Case III: the population is not normal AND the sample size is small
(i.e., n < 30)
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-12
????
4
The t Distribution
 The t distribution is a specific type of symmetric bell-shaped
distribution with a lower height and a wider spread than the
standard normal distribution. As the sample size becomes
larger, the t distribution approaches the standard normal
distribution. The t distribution has only one parameter, called
the degrees of freedom (df). The mean of the t distribution is
equal to 0 and its standard deviation is .
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-13
) 2 /( ÷ df df
Problems Involving the t Distribution
 How to find probability given t values?
 Table V, tdist(t,df,tails) in Excel, and tcdf(t
1
,t
2
,df) in TI-83
 How to find t values given probability?
 Table V, or tinv(prob-of-two-small-tails,df) in Excel.
 Example 8-4
 Find the value of t for 16 degrees of freedom and .05 area in the right
tail of a t distribution curve.
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-14
Table V only works on right tail
t Distribution Table V
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-15
 Fundamental idea
 Two end points and a probability (confidence)
 Probability comes from a known distribution
 Sample mean’s distribution
 Steps to obtain confidence interval on μ
 Point estimate -----------------
 Critical value ------------------- t from the distribution of T
n-1
 Margin of error
 Confidence interval ----------- ± E
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-16
2
1
1
~ ( , ) ~ (0,1) ~
/ /
Find such that ( ) 0.95 (or in general 1 )
Solve for from
/
n
n
X X
X N N T
n n s n
z P t T t
X
t t
s n
µ o
µ
µ
o
µ
µ
o
÷
÷
÷ ÷
¬ ¬
÷ s s = ÷
÷
÷ s s
X
s
E t
n
= ×
X
Confidence Interval for μ Using the t Distribution
5
Example 8-5
 Dr. Moore wanted to estimate the mean cholesterol level for
all adult men living in Hartford. He took a sample of 25
adult men from Hartford and found that the mean cholesterol
level for this sample is 186 mg/dL with a standard deviation
of 12 mg/dL. Assume that the cholesterol levels for all adult
men in Hartford are (approximately) normally distributed.
Construct a 95% confidence interval for the population mean
μ.
 Solution
 Summary statistics: n = 25, = 186, s = 12 & α = 0.05
 Point estimate: 186
 Critical value of 95%: 2.064 (from T
24
)
 Margin of error: 2.064 x 12/5 = 4.95
 95% C. I. on μ is 186 ± 4.95
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-17
X
Example 8-6
 Sixty-four randomly selected adults who buy books for
general reading were asked how much they usually spend on
books per year. The sample produced a mean of \$1450 and
a standard deviation of \$300 for such annual expenses.
Determine a 99% confidence interval for the corresponding
population mean.
 Solution
 Summary statistics: n = 64, = 1450, s = 300 & α = 0.01
 Point estimate: 1450
 Critical value: z = 2.58 (or as book t = 2.656)
 Margin or error: 2.58 x 300/8 = 96.75 (or 99.6 using t = 2.656)
 99% C. I. on μ is 1450 ± 96.75 (99.60)
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-18
X
8.5 Estimation of Population Proportion p
 Point estimate of population proportion p:
 Sampling distribution of
 Approximate normal distribution with
 Mean = p
 Standard deviation =
 Estimate of the standard deviation
 Idea of confidence interval
 Steps for confidence interval on p
 Point estimate
 Critical value z from Z distribution
 Margin of error E = z
 C. I. interval on p
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-19
ˆ
x
p
n
=
ˆ
x
p
n
=
pq
n ˆ ˆ pq
n
ˆ ˆ ˆ
ˆ ( , ) (0,1)
ˆ ˆ
pq p p
p N p N
n pq
n
÷
~ ¬ ~
ˆ
x
p
n
=
ˆ ˆ pq
n
ˆ p E ±
Example 8-7
 According to a survey by Pew Research Center in June 2009,
44% of people aged 18 to 29 years said that religion is very
important to them. Suppose this result is based on a sample
of 1000 people aged 18 to 29 years.
 What is the point estimate of the population proportion?
 Find, with a 99% confidence level, the percentage of all people aged
18 to 29 years who will say that religion is very important to them.
What is the margin of error of this estimate?
 Solution
 Summary statistics: n = 1000, = 0.44
 Point estimate 0.44
 Critical value of 99%: 2.58
 Margin of error: 2.58 x √(0.44x0.56/1000) =0.04
 99% C. I. on p: 0.44 ± 0.04
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-20

6
Example 8-8
 According to a Harris Interactive survey of 2401 adults
conducted in April 2009, 25% of adults do not drink alcohol.
Construct a 97% confidence interval for the corresponding
population proportion.
 Solution
 Summary statistics: n = 2401, = 0.25
 Point estimate: 0.25
 Critical value: 2.17
 Margin of error: 2.17 x √(0.25x0.75/2401) = 0.019
 97% C. I. on p: 0.25 ± 0.019
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-21

Sample Size for Estimation of p
 Given the confidence level and the values of , the sample
size that will produce a predetermined maximum of error E of
the confidence interval estimate of p is
 In case the value of is not known, we often make the most
conservation estimate of the sample size n by using = 0.5
 Or take a preliminary sample (of arbitrarily determined size) and
calculate from this sample. Then use this value to find n.
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-22

=
2
2
ˆ ˆ
E
z p q
n

Example 8-9
 Lombard Electronics Company has just installed a new
machine that makes a part that is used in clocks. The
company wants to estimate the proportion of these parts
produced by this machine that are defective. The company
manager wants this estimate to be within .02 of the
population proportion for a 95% confidence level. What is
the most conservative estimate of the sample size that will
limit the maximum error to within .02 of the population
proportion?
 Solution
 95% => z = 1.96
 Take = 0.5
 Therefore, n = 1.96 x 0.5 x 0.5 / 0.02
2
= 2401
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-23

Technology Instruction
 TI – 83 plus / TI – 84
 Data set
 Summary statistics
 Excel
 Data set
 Function Confidence
MATH 3038 - 01 Dr. Yingfu (Frank) Li 8-24