Statistics and Probability

Estimation of Parameters

Nida C. Flores
Teacher III
 Introduction to Estimation of Parameters

Lesson Objectives
 Illustrates the T–Distribution
(M11/12SP- IIIg-2)
 Identifies Percentiles using the T–Table
(M11/12SP-IIIg-5)
 Introduction to Estimation of Parameters

LESSON 1
Introduction to Estimation of Parameters
 Introduction to Estimation of Parameters

Expected Outcomes
 Show the difference between point and
interval estimate
 Determine the components of confidence
interval
 Introduction to Estimation of Parameters

ESTIMATION OF PARAMETERS

Estimation of parameters refers to the process of approximating a parameter based on

statistics. A parameter refers to any measure that describes a population, while statistic
refers to any measure that describes a sample.
When it is impractical to gather data from all the units of a population, it becomes
necessary to make an estimate. In this situation, statistical data are analyzed to make a
good approximate of the parametric data.
Based on certain statistical principles, the basis of estimation of parameters are the
data gathered from a sample, is usually close to the data that would have been gathered
from a population, provided that an appropriate sample size is used, the situation being
studied is fully understood, and the techniques are done accurately
 Introduction to Estimation of Parameters
Types of Estimates

A point estimate is a specific numerical value of a population parameter. The sample

mean 𝑥̅ is the best point estimate of the population mean. Sample median and sample mode
may also serve as a point estimate of the population mean
An interval estimate is a range of values that may contain the parameter of a
population.
Characteristics of a Good Estimator
An estimator refers to a measurement taken from a sample gathered to represent the
population. The mean, median or mode of the sample can be used as estimators.
Here are the characteristics of a good estimator:
1. The estimator should be unbiased. This means that the mean of the distribution of the
possible values of the estimator is equal to the parameter being observed.
2. The estimator should be consistent. This means that the value of the stimator must approach
the parameter being studied as the sample size increases.
3. The estimator should be relatively efficient. This means that the estimator should have a
small variance
 Introduction to Estimation of Parameters
Types of Estimates

A point estimate is a specific numerical value of a population parameter. The sample

mean 𝑥̅ is the best point estimate of the population mean. Sample median and sample mode
may also serve as a point estimate of the population mean
An interval estimate is a range of values that may contain the parameter of a
population.
Characteristics of a Good Estimator
An estimator refers to a measurement taken from a sample gathered to represent the
population. The mean, median or mode of the sample can be used as estimators.
Here are the characteristics of a good estimator:
1. The estimator should be unbiased. This means that the mean of the distribution of the
possible values of the estimator is equal to the parameter being observed.
2. The estimator should be consistent. This means that the value of the stimator must approach
the parameter being studied as the sample size increases.
3. The estimator should be relatively efficient. This means that the estimator should have a
small variance
 Introduction to Estimation of Parameters
Interval Estimate and Confidence Interval

It is not always a good idea to rely on a single value in estimating parameters, because
the data from a sample can always vary from the population measure as we have learned in
sampling distribution. As such it is better if a range of values is used in estimating parameter.

The range of values that intends to estimate a parameter is called interval estimate. An
interval estimate is better form of estimate because it considers the data from the sample,
might deviate from the actual parameter. The only possibilities are that either interval
estimate may or may not contain the population measure

A form of interval estimate that is mostly used in statistics is called a confidence

interval. A confidence interval is an interval estimate constructed from a point estimate and a
margin of error. The margin of error refers to the maximum estimate of how far the
parameter could be from the point estimate based on an established confidence level
 Introduction to Estimation of Parameters
Example.

A confidence interval used to estimate a population mean was determined to be 67 . The

sample mean is equal to 67. This means that the confidence interval estimates the population
mean to be , at most four units away from 67. Therefore the, the confidence interval 67 4
estimates that the range of values from 63 to 71 contains the population mean.

The margin of error is determined by establishing a confidence level. Confidence level ,

also known as degree of confidence, tells us how sure we would like to be that the confidence
interval contains the parameter it estimates. It is denoted by (1-). If is set to be 0.05, then the
confidence level is 0.95 or 95%
Example.
If a confidence level is established to be 0.95 or 95% , it means that we are 95 percent
confident that the interval estimate contains the parameter being estimated. Another way of
saying is that , if an interval estimate is constructed100 times, 95% of those will contain the
parameter.
 Introduction to Estimation of Parameters
Confidence Levels

The confidence levels that are most often used are 90%, 95%, and 00%. With all things
being equal, as the confidence level is increased, the margin of error must be increased too.
This means that to be more confident with our estimate, we compensate by having a wider
interval estimate. If we aim for a narrower interval estimate, we must establish a lower
confidence level
 Estimation of Population Mean Using the CLT

LESSON 2
Estimation of Population Mean Using the Central Limit
Theorem
 Estimation of Population Mean Using the CLT

Expected Outcomes
 Identify the confidence interval for estimating means using the
central limit theorem
 Identify the appropriate interval estimator when the sample
size is greater than 30 and the population standard deviation is
assumed to be known
 Solve problems involving determining the confidence interval
of the population mean.
 Compute for the sample size required for a length of a
confidence interval
 Estimation of Population Mean Using the CLT
Constructing a Confidence Interval Using the Standard Normal Distribution

The following are the condition for constructing the confidence interval using the
CLT(standard normal distribution):
1. The sample size is greater than30,if not , the population resembles a normal distribution.
2. The population standard deviation is known.
3. The population size is infinite or very large ( infinite population correction factor is very
close to1)
We can estimate the population mean by constructing a confidence interval using the
CLT. It gives a z- score equal to the equation,

Where,
z= random variable in a normal distribution
=sample mean

n = sample size
 Estimation of Population Mean Using the CLT
Constructing a Confidence Interval Using the Standard Normal Distribution

The Standard Normal Distribution Table can be used to determine the areas bound by
two z-scores. This area indicates the probability that the random variable z will fall within the
boundaries. In estimation we use the word confidence instead of probability. The general rule
is that, “confidence” is used once the data have been gathered, and “probability” is used if the
data have not been gathered yet .
From this idea we can determine with a certain level of confidence, the margin of error in
a confidence interval. We can describe the margin of error as the maximum value of − μ and
can be computed by deriving the equation,

− μ = ()
 Estimation of Population Mean Using the CLT
Constructing a Confidence Interval Using the Standard Normal Distribution

The area of the confidence interval is equal to the confidence level. The Z- scores bound
the confidence interval. Those serve as the maximum limits as to how far the z-score will
deviate from the population mean
 Estimation of Population Mean Using the CLT

Constructing a Confidence Interval Using the Standard Normal Distribution

If the area bound by the confidence interval is equal to(1- and the entire area below the
curve is equal to 1, then the area outside the confidence interval is equal to . Let us refer to
these areas as the tails. The area bound by each tail is equal to . This is important in
determining the z-scores that bound the confidence interval. Those two z-scores are the
boundaries of the confidence interval to be .
 Estimation of Population Mean Using the CLT
Constructing a Confidence Interval Using the Standard Normal Distribution

The z- scores that bound the confidence interval are related to margin of error. They are
the z-scores where () is greatest. Once established, we can solved the margin of error using
the equation,
− μ = ()

E= (

Where,
E = margin of error
= z- score that bounds the confidence interval
population standard deviation
n= sample size
Note: As the standard deviation increases, the marginof error widensand vice versa.
Howeverasthesample size increases, the margin oferror narrowsand vice versa
 Estimation of Population Mean Using the CLT
Confidence Interval of Population Mean Using Central Limit Theorem

,
μ = [ · ()]
Where,
= sample mean
= z- score that bounds the confidence interval
population standard deviation
n= sample size

Another way of stating the confidence interval is,

[ · ()]
This means that you are confident up to a certain level of confidence that the population
mean will be within the interval ranging from [ · ()]
 Estimation of Population Mean Using the CLT

The confidence level is usually set at the following levels with confidence coefficients;

a. 0. 99 or 99% confidence level,

b. 0.95 or 95% confidence level,
c. 0.90 or 90% confidence level, = 1.65

The confidence level cannot be set at 100 percent. Recall that the entire normal
distribution stretches towards infinity. Therefore , setting the confidence levelat 100 percent
will give a margin of error equal to infinity.
 Estimation of Population Mean Using the CLT

The steps in determining the confidence interval of a population mean if the

sample size is large are as follows:
1. Determine the sample mean.
2. Establish the confidence level. Given the confidence level, solve for
3. Look for the z-score corresponding to the area
4. Solve for the marginof error
5. Write the confidence interval in the form, [ · ()]
 Estimation of Population Mean Using the CLT
Example1.

A team of educators would like to know the average time it takes the Grade 4 students in an
entire province to finish a certain mathematical problem. To do this , 60 students were asked to
solve the problem and the average time they got is 88.00 seconds. If the population standard
deviation is 4.00 seconds, what is the margin of error for , if
Solution.
Given the , the confidence level is then set up at 0.95 or 95%. Therefore, is equal to
1.96.The confidence interval

μ = [ · ()]

μ = [1.96· ()]

μ=

or, 86.99 s

Answer. At 95 % confidence, the confidence interval is 88.00

 Estimation of Population Mean Using the CLT

If the population standard is not yet determined beforehand, the sample standard
deviation s can be used to estimate as long as the sample size is greater than30.
Example2.
The president of a university would like to know the average commuting time of the
students when travelling to school. To do this, he asked a researcher to gather data
from100 students . The sample mean was found to be 34.05 minutes and the standard
deviation to be 9.50minutes. At a confidence level of 90 percent, what is the confidence
interval that can estimate the population mean?
Solution.
Given the confidence level of 95% or 0.90. is equal to 0.10. Therefore, is equal to
1.65.The confidence interval

μ = [ · ()]
 Estimation of Population Mean Using the CLT

Solution.
Given the confidence level of 95% or 0.90. is equal to 0.10. Therefore, is equal to
1.65.The confidence interval

μ = [ · ()]

μ = [1.65· ()]min

μ = 1.57min

or, 34.05- 1.57 min

Answer. At 90 % confidence level, the average commuting time of the students is estimated at 34.05
 Estimation of Population Mean Using the CLT
Your Turn!

A researcher would like to know the average monthly salary of teachers in a province. To do this,
he made a sample of 150 teachers and found their average salary If the average monthly salary of
all the teachers in the province has a standard deviation of P 1, 500.00 determine the confidence
interval of the population mean if

μ = [ · ()]
Solving for the Sample Size Given the Margin of Error and Level of Confidence

E = · ()

( )
2
𝑍𝛼
•𝜎
2
𝑛=
𝐸

Where,

n = sample size
=corresponding z-score of the confidence level (1-
standard deviation of the population
E = margin of error
 Estimation of Population Mean Using the CLT
Example 1.

A researcher would like to know the average time it takes the students of Bahay Pare Elementary
School to read a 5,000- word essay. He would like to have a margin of error of 1.00 minute and a
confidence level of 95 percent. The population standard deviation was already determined to be
5.45 minutes. How many students should the researcher consider for his sample

( )
2
𝑍𝛼
•𝜎
2
𝑛=
𝐸

114.05
 Estimation of Population Mean Using the CLT
Your Turn!

A researcher wants to know the typing speed of the students of a large TESDA school. The
population standard deviation is previously knownto be 2.3 minutes. If the researcher wants a
margin of error at 0.50 minute and a . What should be the sample size?

( )
2
𝑍𝛼
•𝜎
2
𝑛=
𝐸
 Estimating the Mean Using the t-Distribution

LESSON 3
Estimating the Mean Using the T- Distribution
 Estimating the Mean Using the t-Distribution

Expected Outcomes
 Determine the properties of a t- distribution
 Use t-statistics to determine confidence interval estimate of
the mean.
 Estimating the Mean Using the t-Distribution
T-Distribution

The population mean can be estimated using t- statistic if the sample size
is not large or n is less than 30 or the sample standard deviation is unknown.
A t- statistic, also called t- score , is a random variable that assumes the value,

or

Where,
t = t- score
= sample mean
population mean
= sample standard deviation
n= sample size
E= margin of error
 Estimating the Mean Using the t-Distribution

Characteristics of T-distribution

1. It is composed of bell-shaped curves. Each curve represents a degree of freedom.

2. Its distribution is symmetric about the mean.
3. The distribution stretches towards positive and negative infinities, and it never has a
value below zero.
4. The area under the curve bound by two-t-scores is equal to the probability that the t-
score will fall within that area.
5. The total area under any of the curves is equal to 1
6. The variance of the distribution is greater than one unit
7. As the sample increases, the t-distribution approaches the standard normal distribution
 Estimating the Mean Using the t-Distribution

Degree of Freedom

The distribution of t- scores is composed of a family of curves. Each curve

represents a degree of freedom (d.f.). The degree of freedom refers to the number of
observation s in the sample that can take on different values. This varies depending on
what kind of data is being gathered from the sample. If the sample mean is being taken , The
degree of freedom is equal to (n-1). The t- distribution resembles a standard normal
distribution as the degree of freedom exceeds 29 or 30. When the sample size is equal or
greater than30, the standard normal distribution can be used to approximate the t-
distribution .
Degree of Freedom
d. f
Where,
= degree of freedom
n= sample size
 Estimating the Mean Using the t-Distribution

Understanding T-Distribution with the Degree of Freedom

To determine the degree of freedom, you have to know how many observations in
a sample can vary once the statistic is established.
For example , if the sample size is composed of four students and the mean of their test
score is equal to15. The degree of freedom refers to t how many units of that sample can
vary while maintaining the sample mean of 15. Three of the those four students can have
varying values. If the first student got 6, the second student can still get any value and the
sample mean can still be 15. If the second student got 10, the third student’s score can
assume any value and the sample mean can still be 15.However if the third student’s score
was also determined and it turned out to be 21, the fourth student has to get 23 to have
sample mean of 15. You can assign any value for the scores of the first three students, but
you will realize that for any assigned value, the fourth student cannot just assume any value
and still get a sample mean of 15.
 Estimating the Mean Using the t-Distribution
Example 1.

If a research has a sample size of 28, what is the degree of freedom for getting the
sample mean?
 Estimating the Mean Using the t-Distribution

Estimating Population Mean Using t-Distribution

Solving for Margin of Error Using the t-distribution

or

•
Where,
E = margin of error
= t-score that bounds the confidence level
S = sample standard deviation
n = sample size
 Estimating the Mean Using the t-Distribution

Estimating Population Mean Using t-Distribution

Solving for Confidence Interval Using the t-distribution

Where,
=population mean
= sample mean
= t-score that bounds the confidence level
S = sample standard deviation
n = sample size
 Estimating the Mean Using the t-Distribution
Example1.

A team of educators would like to know the average time it takes the Grade 4 students in an
entire province to finish a certain mathematical problem. To do this , 12 students were asked to
solve the problem and the average time they got is 88.00 seconds. If the population standard
deviation is 4.00 seconds, what is the margin of error for , if
Solution.
Given that n is less than 30 , the confidence level is then set up at 0.95 or 95%.
Therefore, is equal to 2.201( using t-Table).The confidence interval

μ = [ · ()]

μ = [2.201.96· ()]

μ=

or, 85.46 s

Answer. At 95 % confidence level,

 Estimating the Mean Using the t-Distribution
Your Turn!.

A farmer would like to know the average mass of his one-month-old pigs. To do this, he measured
the mass of 15 pigs and found them to be 15.25 kilograms on average. The sample standard
deviation is 2.12 kilograms. At a confidence level of 95 percent, determine the confidence interval
that will estimate the mean.

Solution.

μ = [ · ()]
 Estimating the Mean Using the t-Distribution

t-statistic vs z- statistic

A common source of confusion in determining the confidence interval to estimate the

population mean is the use of proper estimator. The z- statistic and t- statistic can be used as
estimators depending on the situation

If the sample size is greater than30and the population standard deviation is given, the
z- statistic can be used. It can also be used if the sample size is less than 30 but the population
is normally distributed. A t- statistic however , is used when the sample size is less than30.
 Estimation of Population Proportion Using Central Limit Theorem

LESSON 4
Solving for the Confidence Interval Estimate of a Population
Proportion Using the CLT
 Estimation of Population Proportion Using Central Limit Theorem

Expected Outcomes
 Compute for the confidence interval of the population
proportion
 Solve problems involving confidence interval estimates of a
population proportion
 Draw conclusions about the population proportion based onits
confidence interval estimate
 Solve for the sample size required for the length of confidence
interval
 Estimation of Population Proportion Using Central Limit Theorem

Case Study

The contractor that makes armchairs for a university would like to know the proportion of the student
population that is left-handed. To do this, he randomly chose a sample of 50 students and found out that six of
them are left-handed. Find the confidence interval of the population proportion thatis left-handed if the
confidence level is at 90 percent.
 Estimation of Population Proportion Using Central Limit Theorem

Sample Proportion
The population proportion, denoted as p, refers to the proportion of the population that has the
characteristic being sought. Proportion is more convenient to express as fraction when involving proportions in
solutions, and use percent for final answers.
The point estimate for population proportion is called sample proportion. The sample proportion,
denoted as is solved by dividing the number of sample that has the characteristic being looked for
divided by the sample size.

Solving for Sample Proportion

=
Where,
= sample proportion
X= number of samples with the characteristic being looked for
n = sample size
 Estimation of Population Proportion Using Central Limit Theorem

Sample Proportion
Another quantity of interest is .This quantity refers to the proportion of the sample that
does not have the characteristic being looked for.

Solving for

Also, can be solved if the sample proportion is given

= 1-
 Estimation of Population Proportion Using Central Limit Theorem
Example1.

A sample of 34 parents was asked to know the proportion of them that has a child
studying in college.It was found out that 12 of them have a child in college Solve for
and

Solution.
Given:
X= 12 =
n= 34
= 0.35 or 35%

= 1-

= 1- 0.35
= 0.65 or 65%
 Estimation of Population Proportion Using Central Limit Theorem
Your Turn!

A sample of 56 journalists was asked to determine how many of them focuses on

sports. It was determined that 14 of them do. Solve for the sample proportion
 Estimation of Population Proportion Using Central Limit Theorem

Estimating Population Proportion Using CLT

The population proportion can be estimated by a confidence interval using z- statistic. The population
proportion is also referred as the true proportion denoted by p
Conditions to be met when estimating population proportion using the z- statistic:
1. The sample size is greater than 30
2. The conditions for binomial distributions are met.

Solving for Margin of Error Population Proportion Using CentralLimit Theorem

√
^𝑞
𝑝 ^
𝐸=𝑍 𝛼
2
𝑛
Where,
E = margin of error
= z- score for the confidence level
= sample proportion

n = sample size
 Estimation of Population Proportion Using Central Limit Theorem

Estimating Population Proportion Using CLT

Solving for Confidence Interval of Population Proportion
p

or

Wh
= population proportion
 Estimation of Population Proportion Using Central Limit Theorem
Example2.

You would like to determine the proportion of the customers of a shop who subscribe to their buy
one –take one promo. Out of 90 customers, you found out that 34 of them do. At a confidence level
of 95%, solve for the confidence interval estimate of the true proportion of the customers who
subscribe to the buy one- take one promo.
Solution. p
Answer. At a confidence level of
Given: n=90 95%, the true proportion of
= = 0.38 customers that subscribe to the
p
= 1- 0.38 = 0.62 buy one- take one promo is
p within 28 to 48 percent

or
0.38
. − 0.1<𝑝<0.38+0.1
0.28

28%
 Estimation of Population Proportion Using Central Limit Theorem
Your Turn!

A store manager would like to determine the proportion of his customers who felt underwhelmed
by their customer service. Among 50 customers that he asked, six of them said they felt
underwhelmed. What is the true proportion of his customers who felt they were underserved?
 Estimation of Population Proportion Using Central Limit Theorem

Estimating Population Proportion Using CLT

Determining the Sample Size for Estimating Proportions

√
^𝑞
𝑝 ^
𝐸=𝑍 𝛼
2
𝑛

( )
2
𝑍𝛼
^𝑞
𝑛=𝑝 ^• 2
𝐸

Where,
E = margin of error
= z- score for the confidence level
= sample proportion

n = sample size
 Estimation of Population Proportion Using Central Limit Theorem
Example1.

A researcher would like to know the proportion of doctors who live in a condominium. A previous
study determined this to be o.3. At a confidence level of 955 what should the sample size be if the
margin of error is set to be at 5%

Solution. =1-
Given: =1-0.3 = 0.7

( )
2
Confidence Level =95% 𝑍𝛼
= 0.3 ^𝑞
𝑛=𝑝 ^• 2
𝐸
= 5% or 0.05

( )
2
1.96
𝑛=( 0.3)(0. 7) •
0.05
. n= 322.7 or 323
 Estimation of Population Proportion Using Central Limit Theorem
Your Turn!
You would like to determine the proportion of students in your school who are left- handed. If you
would like to have a margin of error of 0.02or 2 percent, what should be your sample size?