CHAPTER 9:

Estimation From
Sampling Data
Business Mathematics 41
Presented by: John Melvin V. Baranda
 Learning Objectives
 Explain the difference between a point estimate and an interval estimate for a population
parameter.

 Use the standard normal distribution in constructing a confidence interval for a population
mean or proportion

 Use the t distribution in constructing a confidence interval for a population mean.

 Decide whether the standard normal distribution or the t distribution should be used in con
structing a given confidence interval.

 Determine how large a simple random sample must be in order to estimate a population
mean or proportion at specified levels of accuracy and confidence.
 Introduction
 Point estimate is a single number that estimates the exact value of the population pa
rameter of interest (e.g. 𝜇 𝑜𝑟 𝜋).

 Interval estimate includes a range of possible values that are likely to include the
actual population parameter.

 When the interval estimate is associated with a degree of confidence that it actually
includes the population parameter, it is referred to as a confidence interval.
 Point Estimate
 An important consideration in choosing a sample statistic as a point estimate of the value of a pop
ulation parameter is that the sample statistic be an unbiased estimator.

 An estimator is unbiased if the expected value of the sample statistic is the same as the actual
value of the population parameter it is intended to estimate.

 Three important point estimators introduced in the chapter are those for a population mean (𝜇), a popula
tion variance (𝜎 2 ), and a population proportion (𝜋).
 Interval Estimate
 Interval Estimate - a range of values within the actual
value of the population parameter may fall.

 Interval Limits - the lower and upper values of the inter

val estimate.

 Confidence Interval - An interval estimate for which the

re is a specified degree of certainty that the actual value
of the population parameter will fall within the interval.

 Confidence Coefficient - For a confidence interval, the p

roportion of such intervals that would include the popul
ation parameter if the process leading to the interval we
re repeated a great many times.
 Interval Estimate
 Confidence Level - Like the confidence coefficient, this
expresses the degree of certainty that an interval will in
clude the actual value of the population parameter, but
it is stated as a percentage. For example, a 0.95 confide
nce coefficient is equivalent to a 95% confidence level.

 Accuracy - The difference between the observed sample

statistic and the actual value of the population paramet
er being estimated. This may also be referred to as esti
mation error or sampling error.
 Interval Estimates
Example 1 “In our simple random sample of 2000 households, we found the average income to be
𝑥ҧ = $65,000, with a standard deviation, 𝑠 =$12,000. Based on these data, we have 95% c
onfidence that the population mean is somewhere between $64,474 and $65,526.”

Solution:
 Point Estimate of 𝜇 $65, 000
 Point Estimate of 𝜎 $12, 000
 Interval Estimate of 𝜇 $64, 474 to $65, 526
 Lower and upper interval limits for 𝜇 $64, 474 and $65, 526
 Confidence Coefficient 0.95
 Confidence Level 95%
 Accuracy For 95% of such intervals, the sample mean would not differ from the actu
al population mean by more than $526.
Overview of Methods
 Confidence Interval for the Mean: 𝝈 is known

 This application assumes that either (1) the underlying population is normally distributed or (2)
the sample size is n ≥ 30. For example, 𝑧0.025 would be 1.96. The most commonly used confidence
intervals and the corresponding z values typically used to obtain them are shown below:
 z-interval, Mean
Example 2 From past experience, the population standard deviation of rod diameters produced by a machine has
been found to be 𝜎 =0.053 inches. For a simple random sample of n=30 rods, the average diameter is
found to be 𝑥ҧ =1.400 inches. What Is the 95% Confidence Interval for the Population Mean, 𝜇?
𝜎 0.053
𝑥ҧ ± 𝑧 = 1.400 ± 1.96 or between 1.381 and 1.419 inches
𝑛 30
 z-interval, Mean
Example 3 A simple random sample of 25 has been collected from a normally distribute
d population for which it is known that 𝜎 =17.0. The sample mean has been
calculated as 342.0, and the sample standard deviation is s=14.9. Construct a
nd interpret the 95% and 99% confidence intervals for the population mean.
 The Student’s t Distribution

ҧ
𝑥−𝜇
 The t distribution is the probability distribution for the random variable 𝑡 = .
𝑠/ 𝑛

 It has a mean of zero, but its shape is determined by what is called the number of degrees of free
dom (df).

 For confidence interval applications, the specific member of the family is determined by 𝑑𝑓 = 𝑛 −
1.

 The term degrees of freedom refers to the number of values that remain free to vary once some in
formation about them is already known.

 For example, if four items have a mean of 10.0, and three of these items are known to have value
s of 8, 12, and 7, there is no choice but for the fourth item to have a value of 13. In effect, one degr
ee of freedom has been lost.
 t Distribution Table

 Using the t Distribution Table. A table for t values that correspond to selected areas beneath the
t distribution appears on the pages immediately preceding the back cover. A portion of the table i
s reproduced as Table 9.2. In general, it is used in the same way as the standard normal table, bu
t there are two exceptions: (1) the areas provided are for the right tail only, and (2) it is necessary
to refer to the appropriate degrees of freedom (df) row in finding the appropriate t value.

Example

• For a Sample Size of n = 15, What t Values Would Correspond to an Area Centered at t 5 0 and Ha
ving an Area beneath the Curve of 95%?

• For a Sample Size of n = 99, What t Values Would Correspond to an Area Centered at t 5 0 and Ha
ving an Area beneath the Curve of 90%?
 Confidence Interval Using t Distribution
 The appropriate t value is used instead of z, and s replaces 𝜎.

 The t distribution assumes the underlying population is approximately normally distributed, but this assumption
is important only when the sample is small—that is, n , 30.
 t-interval, Mean
Example 4 A simple random sample of n=90 manufacturing e
mployees has been selected from those working t
hroughout a state. The average number of overtim
e hours worked last week was 𝑥ҧ =8.46 hours, with
a sample standard deviation of s=3.61 hours. Wha
t Is the 98% Confidence Interval for the Populatio
n Mean, 𝜇?

𝑠 3.61
𝑥ҧ ± 𝑡 = 8.46 ± 2.369 or between 7.56 and 9.36 hours
𝑛 90
 t-interval, Mean
Example 5 Given the following observations in a simple random sample from a population that is ap
proximately normally distributed, construct and interpret the 90% and 95% confidence int
ervals for the mean:

67 79 71 98 74 70 59 102 92 96
 Confidence Interval Estimates for the
Population Proportion
 Determining a confidence interval estimate for the population proportion requires that we use the sample proport
ion (p) for two purposes: (1) as a point estimate of the (unknown) population proportion,𝜋, and (2) in combination
with the sample size (n) in estimating the standard error of the sampling distribution of the sample proportion for
samples of this size.
 z-interval, Proportion
Example 6 In an April 2007 NBC News/Wall Street Journal poll, 1008
adults were randomly sampled from across the United Stat
es. In response to the question, “All in all, do you think that
things in the nation are generally headed in the right direc
tion, or do you feel that things are off on the wrong track?”
22% responded “headed in the right direction.”

What Is the 95% Confidence Interval for the Population Pr

oportion Who Would Have Answered “Headed in the Right
Direction” to the Question Posed?

𝑝(1 − 𝑝) 0.22(1 − 0.22)

𝑝±𝑧 = 0.22 ± 1.96 = 0.194 𝑡𝑜 0.246
𝑛 1008
 z-interval, Proportion
Example 7 A pharmaceutical company found that 46% of 1000 U.S. adults surveyed knew neither their blood pre
ssure nor their cholesterol level. Assuming the persons surveyed to be a simple random sample of U.S
. adults, construct a 95% confidence interval for 𝜋 = the population proportion of U.S. adults who wo
uld have given the same answer if a census had been taken instead of a survey
Estimating the Population Mean
 Sample Size, Estimating a Mean
Example 8 A state politician would like to determine the average amount earned during summer employment by
state teenagers during the past summer’s vacation period. She wants to have 95% confidence that the
sample mean is within $50 of the actual population mean. Based on past studies, she has estimated t
he population standard deviation to be 𝜎 =$400. What Sample Size Is Necessary to Have 95% Confid
ence That 𝑥ҧ will be within $50 of the Actual Population Mean?

𝑧 2 ∙ 𝜎 2 1.962 ∙ 4002
𝑛= = = 245.9 𝑝𝑒𝑟𝑠𝑜𝑛𝑠, 𝑟𝑜𝑢𝑛𝑑𝑒𝑑 𝑢𝑝 246
𝑒2 502
Estimating the Population Proportion
Example 9
Sample Size, Estimating a Population Proportion
A tourist agency researcher would like to determine the proportion of U.S. adults who have ever vacat
ioned in Mexico and wishes to be 95% confident that the sampling error will be no more than 0.03 (3
percentage points). Assuming the Researcher Has No Idea Regarding the Actual Value of the Populati
on Proportion, What Sample Size Is Necessary to Have 95% Confidence That the Sample Proportion
Will Be within 0.03 (3 Percentage Points) of the Actual Population Proportion?

𝑧 2 ∙ 𝑝(1 − 𝑝) 1.962 ∙ (0.5)(1 − 0.5)

𝑛= = = 1067.1 𝑝𝑒𝑟𝑠𝑜𝑛𝑠, 𝑟𝑜𝑢𝑛𝑑𝑒𝑑 𝑢𝑝 1068
𝑒2 0.032

If the Researcher Believes the Population Proportion Is No More Than 0.3, and Uses p 5 0.3 as the Estimate, What
Sample Size Will Be Necessary?

𝑧 2 ∙ 𝑝(1 − 𝑝) 1.962 ∙ (0.3)(1 − 0.3)

𝑛= = = 896.4 𝑝𝑒𝑟𝑠𝑜𝑛𝑠, 𝑟𝑜𝑢𝑛𝑑𝑒𝑑 𝑢𝑝 897
𝑒2 0.032