Professional Documents
Culture Documents
ST130 - Chapter 8
ST130 - Chapter 8
Objectives
After completing this chapter, you should be able to:
1. Find the confidence interval for the mean when the population standard deviation is known.
2. Determine the minimum sample size for finding a confidence interval of mean.
3. Find the confidence interval for the mean when the population standard deviation is unknown.
4. Find the confidence interval for the population proportion.
5. Determine the minimum sample size for finding a confidence interval of proportion.
8.1 Introduction
As part of inferential statistics, we need to determine the value of the population parameters. This is not
possible since the population is large, so statisticians have to estimate the value of the parameter. An
important aspect on inferential statistics is estimation, which is the process of estimating the true value
of a population parameter from the information derived from a small sample. For instance, the population
mean ( ) can be estimated using the sample mean ( ).
Therefore, in this chapter, we will explain statistical procedures for estimating the population mean and
proportion. Another important question in estimation is that of sample size. How large should the sample
be drawn in order to make an accurate estimate? This question is not easy to answer as it depends on
several factors, such as the accuracy desired and the probability of making a correct estimate. The
problem of determining the sample size for estimating the parameters will also be discussed in this
chapter.
8.2 Estimation
Estimation is the process of estimating the true value(s) of a population parameter from the information
derived from a small sample.
Confidence Interval
A confidence interval is a specific interval estimate of a parameter determined by using the data obtained
from sample and a specific confidence level.
Confidence Level
A Confidence Level of an interval estimate of a parameter is the probability that the interval estimate will
contain the parameter.
Suppose that a 90% confidence interval states that the population mean is greater than 100 and less
than 200. How would you interpret this statement?
SOLUTION
It means that we are 90% confident that the interval contains the true population mean.
8.3 Confidence Intervals and Sample Size for the Mean when is
known
Before constructing the confidence interval for , it is essential to know the following:
Is the distribution of the population normal or not?
Is the population standard deviation known or unknown?
Is the sample size large or small?
Our answers will then determine how to proceed. In this section we are going to construct the confidence
interval of the population mean when is known.
X z 2 X z 2 .
n n
Note:
1. If n < 30 the population should be normally distributed.
2. The values of for some confidence interval are as follows:
For the 99% confidence interval, = 2.58.
For the 95% confidence interval, = 1.96.
For the 90% confidence interval, = 1.65.
However, other values for confidence level could be given, so how do we find the value of Let us
consider the next example.
SOLUTION
Draw a standard normal curve and shade the area 0.98 in the middle. See the graph below.
0.98
Use the standard normal table from the Eton tables to find the value of Lookup 0.49 in the
probability section and read the corresponding z value. Therefore = 2.33.
Note: The value of 0.98 = 0.02 and / 2 = 0/02 / 2 = 0/01. Therefore, the area on the right of
is 0.01.
SOLUTION
Hence one can say with 99% confidence that the average spending per visit at MHCC Bookstore is
between $22.42 and $24.48, based on a sample of 49 customers.
Suppose a registrar of the University of the South Pacific (USP) wishes to estimate the average number
of hours per day of distractions (phone calls, emails, impromptu visits, etc.) experienced by USP lecturers.
A study of random sample of 50 lecturers in USP found that the average distraction time is 1.8 hours per
day and the population standard deviation was 20 minutes. Estimate the true mean population distraction
time for USP lecturers with 90% confidence.
SOLUTION
Hence one can say with 90% confidence that the average distraction time for a USP lecturer is between
1.72 and 1.88 hours per day, based on 50 lecturers.
Sample Size
Quite often, researchers need to know how large the sample is necessary to make an accurate estimate.
One may ask why sample size is so important. The answer to this is that an appropriate sample size is
required for validity. If the sample size is too small, it will not yield valid results. An appropriate sample
size can produce accuracy of results. Moreover, the results from the small sample size will be
questionable. A sample size that is too large will result in wasting money and time.
Where,
is called the margin of error.
A pizza shop owner wishes to find the 95% confidence Interval of the true mean cost of a large plain
pizza. How large should the sample be if she wishes to be accurate to within $0.15? A previous study
showed that the population standard deviation of the price was $0.26.
SOLUTION
Therefore, the minimum sample size should be 12 to estimate the population mean with 95%
A researcher in Fiji wishes to estimate within $300 the true average amount of money Fiji spends on road
repairs each year. The standard deviation is known to be $900. If she wants to be 90% confident, how
large a sample is necessary?
SOLUTION
Therefore, the minimum sample size should be 25 to estimate the population mean with 90% confidence.
SOLUTION
For the 90% confidence interval, = 0.10, thus /2 = 0.05. Since n = 20, d. f. = 20 1 = 19, so look up the
t-distribution table with = 19, 2p = 0.1 and p = 0.05 and we get to be 1.729.
For a group of 20 ST130 students subjected to a stress situation, the mean number of heart beats per
minute was 126, and the standard deviation was 4. Find the 95% confidence interval of the true mean.
Assume the variable is normally distributed.
SOLUTION
Since the population standard deviation, is unknown, we use the t-distribution. For the 95%
Confidence Interval, = 0.05 /2 = 0.025 and the d.f. t-distribution table
from the Eton tables with = 19, 2p = 0.05 and p = 0.025 and we get to be 2.093. Now the 95%
confidence interval is:
44 52 31 48 46 39 47 36 41 56
SOLUTION
B. Similarly, is unknown, so we use the t-distribution. Look up the t-distribution table with = 9,
2p = 0.05 and p = 0.025 and we get to be 2.262. Hence the 95% confidence interval is:
The population proportion, denoted by is the proportion of population units that possess a
characteristic. The population proportion is given by:
Where,
is the number of population units that possess a characteristic
is the population size
q = 1 p, is the proportion of population units that do not possess a characteristic
For example, in the USP assessment meeting, the ST130 lecturer stated that 75% of ST130 students
pass the course last semester. The parameter 65% is a population proportion.
The population proportion, is often unknown, so a sample proportion, denoted as (read p hat) is
used to estimate it. It represents the proportion of sample units that possess a characteristic. The sample
proportion is given by:
In a study, 400 students were interviewed if they own a computer; 352 said that they had computers. Find
SOLUTION
A recent study of 100 people in Fiji found 27 were obese. Find the 95% confidence of the population
proportion of all individuals living in Fiji who are obese.
SOLUTION
Hence, one can be 95% confident that the proportion of people obese in Fiji is between 18.3% and 35.7%.
A survey of 120 female freshmen showed that 18 did not wish to work after marriage. Find the 90%
confidence interval of the true proportion of females who do not work after marriage.
SOLUTION
Hence, we can say with 90% confident that between 9.6% and 20.4% of females do not work after
marriage.
Where,
is called the margin of error
It is believed that 10% of Suva homes have a direct satellite television receiver (SKY Pacific). How large
a sample is necessary to estimate the true population of homes which do with 90% confidence and within
3 percentage points?
SOLUTION
A researcher wishes to estimate the proportion of executives who own a car phone. She wants to be 99%
confident and be accurate within 5% of the true proportion. Find the minimum sample size necessary.
SOLUTION
For 99% confidence interval, In this problem, we have no prior knowledge of and so we
assign and therefore, . Hence,
EXERCISES
2. A recent survey of 8 social networking sites has a mean of 13.1 and a standard deviation of 4.1
million visitors for a specific month. Find the 95% confidence interval of the true mean. Assume that
the variable is normally distributed.
3. If the variance of a national accounting exam is 900, how large a sample is needed to estimate the
true mean score within 5 points and with 99% confidence?
4. The number of unhealthy days based on the AQI (air quality index) for a random sample of
metropolitan areas is shown:
61 12 6 40 27 38 93 5 13 40
A. What is the point estimate of the mean number of unhealthy days all such days?
B. Construct a 98% confidence interval of based on the data.
5. A sample of 30 networking sites for a specific month has a mean of 26.1. Assume the population
standard deviation to be 4.2. Find the 99% confidence interval of the true mean.
6. A recent study indicated that 29% of the 100 women over age 55 in the study were widows. How
large a sample must you take to be 90% confident that the estimate is within 0.05 of the true
proportion of women over age 55 who are widows?
7. A Tongan advertising agency wishes to estimate the proportion of household, which use a particular
brand of washing soap. They decide on the sample size of 500 and find that 157 households use the
product.
A. Construct a 99% confidence interval for proportion.
B. How large should a sample have to be for their interval estimate of proportion to have been in
error by 2%?
8. In a survey of drug use among 995 Suva teenagers, the following results were reported. Estimate
with 90% confidence the proportion of all Suva teenagers who are daily smokers or occasional
smokers.