CONFIDENCE LEVEL

AND SAMPLE SIZE

Statistics and Probability
 Starting Point

ENTRY CARD
A. Fill in the blanks with the correct word or group of words to make meaningful
statements.
1. The round up of 110.12 is .
2. A confidences level is the confidence coefficient expressed as .
3. For a 95% confidence level, the confidence coefficient is .
4. When 99% confidence is selected, a = .
5. The standard deviation of a sampling distribution is called .
 Starting Point
ENTRY CARD
6. The subset of a population of interest is called .
7. When 𝜎is unknown, the approximation is given by
.
B. Evaluate each of the following:
1-0.37 5. (
(2.58)(10) 2
0.10 )
2. (0.35)(0.65)
2
1.96(2.3) (2.58)(3.1) 2
3. 0.06 7. 0.10 )
2
(1.96)(5) (2.58)(10)
4. ( 0.02 ) 6. 0.5 )
 Getting Ready

• How large should a sample be? Many researchers find this a problem. In
fact, many rely on old methods to determine the sample size they need in
their investigations. This problem should be addressed carefully because
needlessly large samples are a waste of important resources. On the other
hand, insufficient sample size may lead to poor results.

Task: Derive the formula in determining sample size.

 In-class Activity
1
• There are two things to remember when we decide on the quality of
the sample size we need: confidence and narrowness of the interval.
The computing formula in determining the sample size is derived from
the formula of the margin of error E where:
𝑛= 𝑧𝑎 2 ( 𝜎
√𝑛
)
• Study the derivation of the computing formula for determining the minimum sample
size when estimation a population mean.
In-class
• Discuss the elements of theActivity
formula.
1 n, the sample size needed in estimating a population mean. To do this,
Let us solve for
we modify the equation by applying the properties of equality.

Formula in Determining the

𝑧𝑎2 ∙ 𝜎
So, 𝐸= Minimum Sample Size Needed
√𝑛
when Estimating the Population
𝐸 𝑛 = 𝑧𝑎 2 ∙ 𝜎 Mean
𝑧𝑎2∙𝜎
𝑛= 2
𝐸
2 𝑧𝑎2 ∙ 𝜎 2 𝑛= ( 𝑧𝑎2 ∙ 𝜎 )
( 𝑛) = ( 𝐸) 𝐸
 Since the value of 𝜎is usually unknown, it can be estimated by the
standard deviation s from a prior sample. Alternatively, we may
approximate the range R of observation in the population and make a
conservative

𝑅
estimate of ≈ . In
4 any case, round up the value of
obtained to ensure that the sample size will be sufficient to achieve
the specified reliability.
 Example 1: FEEDING PROGRAM

In a certain village, Leony wants to estimate the mean weight

μ, in kilograms, all of six-year old children to be included in the
feeding program. She wants to be 99% confident that the
estimate of μ is accurate to within 0.06 kg. suppose from a
previous study, the standard deviation of the weights of the
target population was 0.5 kg, what should the sample size be?
 Solution:

Give the confidence 99%. Then a=1 – 0.99 = 0.01. thus,

𝑧𝑎 2 =2.58.
The phrase “accurate to within 0.06 kg” indicates a narrowed
width of the confidence interval. Thus, the desired error E =
0.06 kg.
Population standard deviation 𝜎=0.5 kg
 …

Substituting the values in the equation for getting sample size, we have:
2
n=(
𝑧𝑎2 ∙ 𝜎 )
𝐸
2.58 (0.5) 2
n =( 0.06 )
2
1.29 )
n =( 0.06
n =(21.5)2
n = 462.25
Round up the value to 463. So, the required sample is 463 six years old
children.
 Example 2: Replicating a Study

Joshua wants to replicate a study where the lowest

observed value is 12.4 while the highest is 12.8. He
wants to estimate the population mean μ to within an
error of 0.025 of its true value. Using 99% confidence
level, find the sample size 𝑛that he needs.
 Solution:

The confidence level is 99%. So, =0.01. Therefore, 𝑧𝑎 2=2.58.

𝑅
The desired error is 0.025. since the range 𝑅=12.8-12.=0.4, then 𝜎= =0.01.
4
Substituting the values in the equation for getting sample size, we have:
2
n=(
𝑧𝑎2 ∙ 𝜎 ) Note:
𝐸
2 When determining sample size, we
n=( 2.58 (1.5)
0.025 ) always round up the resulting value
2
) to the next whole number.
.258
n =( 0.025 2
n=(10.32)
Round up 106.5 to 107. So the required sample size is 107.
n =106.5
 In-class Activity
2:
• Study the derivation of the computing formula in determining the sample size
needed when estimating population proportion.
• Compare this formula with the formula derived In-Cass Activity 1.
To determine sample for finding the confidence interval for a proportion, the
computing formula is also derived from the margin of error.

𝐸= 𝑧 𝑝𝑞
𝑎2 𝑛
 Modifying the proportion, we get:
…
𝐸 = 𝑝𝑞
𝑛
𝑧𝑎2

Next, we square both sides.

𝐸 2
( ) = 𝑝𝑛𝑞
𝑧𝑎2

𝐸2 2 𝑝𝑞
(𝑧𝑎 2 ) = 𝑛

𝑛∙ 𝐸2=𝑝𝑞∙ (𝑧𝑎2 ) 2
Finally, solving for 𝑛:
𝑛= 𝑝𝑞∙ 𝑧 2 𝑧𝑎 2 𝟐
𝑎2
or 𝑛= 𝑝𝑞( )𝐸
Formula in Determining the Minimum Sample Size
Needed when Estimating the Population Proportion
𝑧𝑎2
𝑛 = 𝑝 𝑞( ) 𝟐
𝐸
 In computing the sample size for estimating a population proportion,
there are 2 situations to bear in mind:
1. If some approximation of 𝑝is known, that the value may be used in the formula.
2. If no approximation is known, use 𝑝= 0.5. this value will result in a sample size large
enough to guarantee prediction, given the confidence interval and the error of
estimate. The reason is that when 𝑝and 𝑞are 0.5 each, the product is at maximum,
as shown in the following table.
 𝑝 𝑞 𝑝𝑞
0.1 0.9 0.09
0.2 0.8 0.16
0.3 0.7 0.21
0.4 0.6 0.24
0.5 0.5 0.25
0.6 0.4 0.24
0.7 0.3 0.21
0.8 0.2 0.16
0.9 0.1 0.09
 Determining Sample Size involving Proportion

• Study the following steps in determining sample size dealing with

proportions.

Steps in Solving for Sample Size involving Proportions

Step 1. Determine the confidence level.
Step 2. Determine the confidence coefficient.
 Steps in Solving for Sample Size involving Proportions
Step 3. Determine the error E.
Step 4. Determine 𝑝and 𝑞.
Step 5. Substitute the values in
the formula:
𝑧𝑎2
𝑛 = 𝑝𝑞( 𝐸
)𝟐
Step 6. Round up the resulting value to the nearest whole number.
 Example 1: Detergent X

Mary wants to know, with 95% confidence, the

proportion of households who like to use detergent X. a
previous survey showed that 42% like to use detergent X.
Mary like to be accurate within 2% of the true proportion.
What sample size does Mary need?
 Solution:

The computing formula: 𝑧𝑎2

𝑛 = 𝑝 𝑞( 𝐸 ) 𝟐

The error E is pegged at 2%.

Thus, 𝑞= 1 – 0.42 = 0.58
We substitute these values in the computing formula:

Steps
Solution: Solution
1. Determine the confidence level 95% confidence level
2. Determine the confidence With a 95% confidence level, a
coefficient = 0.5.
So, 𝑧𝑎 2 = 1.96
3. Determine the error E. E = 2% or 0.02
4. Determine 𝑝and 𝑞. The given proportion is 42%.
That is 𝑝= 0.42
So, 𝑞= 1 – 0.42 = 0.58
 Steps Solution
𝑧𝑎2
Solution:
5. Substitute the values in the 𝑛= 𝑝𝑞( )𝟐
formula and compute. 𝑧𝑎2 1.96 2
𝐸
𝑛= 𝑝𝑞( )𝟐 𝑛= (0.42)(0.58)( )
0.02
𝐸
𝑛= (0.42)(0.58)(9604)

6. Round up the resulting value to 𝑛= 2339.53 or 2340

the nearest whole number.

Thus, Mary needs a sample of 2340 respondents.

 Example 2: New School Structure

The Dean of a college wants to use the proportion to

determine the sample size needed to interview regarding their
thoughts about the new school structure. She wants to be
able to assert with a probability 0.95 that her error will be at
most 0.05. similar polls in the past showed that 65% approved
the new structure. How large a sample doesthe Dean need?
Solution:

Steps
Solution: Solution
1. Determine the confidence level 95% confidence level
2. Determine the confidence With a 95% confidence level,
coefficient
𝑧𝑎 2 = 1.96
3. Determine the error E. E = 5% or 0.05
4. Determine 𝑝and 𝑞. The given proportion is 65%. That is
𝑝= 0.0.65
Thus, 𝑞= 1 – 0.65 = 0.35
 Steps Solution
𝑧𝑎2
Solution:
5. Substitute the values in the 𝑛= 𝑝𝑞( )𝟐
formula and compute. 𝑧𝑎2
𝐸
1.96 𝟐
𝑛= 𝑝𝑞( )𝟐 𝑛= 𝑝𝑞( )
0.05
𝐸
𝑛= (0.65)(0.35)(39.2)
𝑛= (0.65)(0.35)
(1536.64)
6. Round up the resulting value to 𝑛= 349.58 or 350
the nearest whole number.
The Dean needs a sample size of 350 respondents.
 Mathematical Journal

Compare the sample sizes that you expect to get for the
following conditions:
1. For 90% confidence, 𝛼= 0.10
2. For 99% confidence, 𝛼= 0.01
 Exercises
𝑧𝑎 2
1. Find ( )𝟐 given each of the following:
𝐸
a. 90% confidence, E= 0.01
b. 90% confidence, E= 0.02
c. 95% confidence, E= 0.15
d. 95% confidence, E= 0.08
e. 99% confidence, E=
0.05
 Exercises

2. Find sample size for each of the following: a. 90%

confidence, E= 0.08, 𝑝 = 0.38
b. 95% confidence, E= 0.05, 𝑝 = 0.38
c. 95% confidence, E= 0.12, 𝑝 = 0.38
d. 99% confidence, E= 0.20, 𝑝 = 0.38
e. 90% confidence, E= 0.25, 𝑝 = 0.38
 Exercises

3. What is the advantage of the sample size formula?

4.What is the effect of the level of confidence on the confidence interval?
5.A particular brand of coffee contains an average of 112 mg of caffeine per
cup with a standard deviation of 20 mg. Joshua wants to investigate the same
to estimate the true population mean caffeine content correct within 5 mg
adopting 95% confidence. How many cups of the same brand of coffee does
he need for a sample?
 Exercises

6.In a gymnasium, a physical exercise has a mean length of 30 minutes with a standard
deviation of 6 minutes. A PE major wants to estimate the true meanlength of the
exercise with maximum error pegged at 0.5 adopting the 95% confidence interval. How
many respondents does he need?
7.In a paper presentation, the average algebraic reason of Grade 8 students in a
mathematical camp was observed to be 80 with a standard
deviation of 4.2. a researcher wants to replicate the study to estimate the true population
mean 𝜇to within .5 maximum error. If they 95% level of confidence is adopted, how many
respondents are needed?
 Look Back and Reflect

You have learned that in determining sample size, you need three
values: confidence level, population standard deviation, and desired
error. Answer the following question and discuss it with a classmate:
• What are the practical reason for learning how to use the
formula in determining sample size?
 CHAPTER SUMMARY

• An estimate is a value that approximates a parameter. It is based on

sample statistics computed from sample data.
• In Statistics, estimation is the process of determining values
parameter.
• The confidence level of an interval estimate of a parameter is the
probability that the interval estimate contains the parameter. parameter.
2
• Variance (𝑠2): 𝑠2 = (𝑋 − 𝑋 )

𝑛− 1
 CHAPTER SUMMARY

• Standard deviation (s): s = (𝑋− 𝑋) 2

𝑛− 1
• An interval estimate, called a confidence interval, is a range of values that is used to
estimate a parameter. This estimate may or may not contain the true parameter
value.
• The confidence level of an interval of a parameter is the probability that the interval
estimate contains the parameter. It describes what percentage of intervals from many
different samples contains the unknown population parameter.
 CHAPTER SUMMARY

• A Four-Step in Computing the Interval Estimate

Step 1. Describe the population parameter of interest (e.g. mean 𝜇) Step 2.
Specify the confidence interval criteria.
a. Check the assumptions.
b. Determine the test statistics to be used.
c. State the level of confidence.
Step 3. Collect and present sample evidence.
a. Collect the sample information.
 CHAPTER SUMMARY

b. Find the point estimate.

Step 4. Determine the confidence interval.
a. Determine the confidence coefficients (for a z or a t-
distribution)
b. Find the maximum error E of the estimate.
c. Find the lower and the upper confidence limits.
d. Describe/interpret the results.
 CHAPTER SUMMARY

• The general expression for the confidence interval when 𝜎is unknown
𝑠
) and the distribution of values is called t-
is given by 𝑋 ±𝑡( 𝑛
distribution.
• The formula for computing the confidence interval using the t-
𝑠 )<𝜇< 𝑋 + (𝑡 𝑠 )
distribution is 𝑋 − t(
𝑛 𝑛
• Proportion is a fraction expressed where the favorable response is in the numerator
and the total number of respondents is in the denominator. The basic operation
involves division. Thus, the result is a decimal value that can be expressed as percent.
 CHAPTER SUMMARY

• The formula for computing a large sample confidence interval for a

population proportion 𝑝is:
𝑝𝑞 𝑝𝑞
𝑝= 𝑧 𝑎 2 <𝑛 𝑝< 𝑝+𝑧 𝑎2 𝑛
• Formula in determining the minimum sample size needed when
estimating the population mean:
2
𝑛= ( 𝑧𝑎2 ∙ 𝜎 )
𝐸
 CHAPTER SUMMARY

• Formula in determining the minimum sample size needed when

estimating the population proportion:
𝟐
𝑛= 𝑝 𝑞( 𝑧
𝑎 2 )
𝐸
 Thank you!
Teacher: Kristiane Joie F. Mico