Professional Documents
Culture Documents
Competencies Covered 1
Key Points 1
Practice 23
Tips 26
Answer Key 28
Review Notes 8
Sample Size Determination
Competencies Covered
In this note, we will discuss the important concepts about the
following competencies:
● identify and compute for the length of the confidence
interval;
● compute for an appropriate sample size using the length of
the interval, and
● solve problems involving sample size determination.
Key Points
Determining the Sample Size
● A sample is a part of a population determined by sampling
procedures where its size is determined as 𝑛.
𝑁
𝑛=
1 + 𝑁𝑒 2
where
𝑛 is the number of samples,
𝑁 is the number of population, and
𝑒 is the margin of error.
𝑁
𝑛=
1 + 𝑁𝑒 2
500
=
1 + 500(0.02)2
500
=
1.2
≈ 417
Thus, the sample size that can be used for the study is 𝑛 = 417.
● We can use the following tables to determine the appropriate sample size given the
size of the population and the precision level needed.
Table 1. Sample size for ±5%, ±7%, and ±10% precision levels where the confidence level
is 95% and 𝑝 = 0.5.
Size of Sample size (𝒏) for precision (𝒆) of
population ±𝟓% ±𝟕% ±𝟏𝟎%
100 81 67 51
125 96 78 56
150 110 86 61
175 122 94 64
200 134 101 67
225 144 107 70
250 154 112 72
275 163 117 74
300 172 121 76
325 180 125 77
350 187 129 78
Table 2. Sample size for ±3%, ±5%, ±7%, and ±10% precision levels where the confidence
level is 95% and 𝑝 = 0.5.
Size of Sample size (𝒏) for precision (𝒆) of
population ±𝟑% ±𝟓% ±𝟕% ±𝟏𝟎%
a
500 222 145 83
a
600 240 152 86
a
700 255 158 88
a
800 267 163 89
a
900 277 166 89
a
1 000 286 169 91
2 000 714 333 185 95
3 000 811 353 191 97
4 000 870 364 194 98
5 000 909 370 196 98
6 000 938 375 197 98
7 000 959 378 198 99
8 000 976 381 199 99
9 000 989 383 200 99
10 000 1 000 385 200 99
15 000 1 034 390 201 99
20 000 1 053 392 204 100
25 000 1 064 394 204 100
50 000 1 087 397 204 100
100 000 1 099 398 204 100
>100 000 1 111 400 204 100
a
The entire population must be sampled.
Example:
Researchers may place the names of the participants on a fishbowl and
determine who will be part of the sample by drawing names on that bowl.
Example:
Given a list of names of students in a certain class, a researcher may select
every 3rd person from that list to determine the participants of his study.
Example:
A group of researchers are conducting a study on the popularity of the
President in the province of Cebu. To do this, they divided the population
according to the different cities in Cebu.
Example:
A teacher may choose to select his students as respondents of his study since
this is more convenient to him rather than going to another school.
○ In judgment sampling, the sampler decides who will or will not be included in
a sample.
○ In quota sampling, the sample composition must reflect the makeup of the
population on some preselected characteristics and it often has a nonrandom
component.
Example:
A group of researchers are conducting a study about how people select
politicians that they will vote. They decided to interview 50 males and 50
females for their study. In this case, they interviewed people in a nearby
shopping mall until they reached their quota.
𝑁−𝑛
𝐹𝑃𝐶 = √
𝑁−1
𝑁−𝑛
𝐹𝑃𝐶 = √
𝑁−1
5 000 − 300
=√
5 000 − 1
4 700
=√
4 999
≈ 0.9696
𝜎 𝜎
𝑥 − 𝑧α/2 ( ) < 𝜇 < 𝑥 + 𝑧α/2 ( )
√𝑛 √𝑛
where
𝑥 is the sample mean,
𝜇 is the population mean,
𝑛 is the sample size,
𝑧α/2 is the confidence coefficient,
𝜎 is the population standard deviation, and
𝜎
𝑧α/2 ( ) is the margin of error.
√𝑛
● If the sample size 𝑛 is at least 5% of the population size 𝑁, we will multiply the margin
of error by the value of the Finite Population Correction Factor. As such, the formula
will be as follows:
𝜎 𝑁−𝑛 𝜎 𝑁−𝑛
𝑥 − 𝑧α/2 ( )⋅√ < 𝜇 < 𝑥 + 𝑧α/2 ( ) ⋅ √
√𝑛 𝑁−1 √𝑛 𝑁−1
Example:
In a certain study, the population size is 𝑁 = 10 000 and the sample size is 𝑛 = 100. The
sample mean is 𝑥 = 42.3 and the population standard deviation is 𝜎 = 12.46.
Determine the length of the confidence interval of the population mean at 95%
confidence level.
Since 𝑛 = 100 and 𝑁 = 10 000, the sample size is 1% of the population. This means that
the FPC is not needed in the formula. Since we are dealing with a 95% confidence level,
the value of the confidence coefficient is 𝑧α/2 = 1.96.
Now that we have all the necessary values, we can now start computing for the length
of the confidence interval of the population mean.
𝑁 = 10 000
𝑛 = 100
𝑥 = 42.3
𝜎 = 12.46
𝑧α/2 = 1.96
𝑠 𝑠
𝑥 −𝑡( ) < 𝜇 < 𝑥 +𝑡( )
√𝑛 √𝑛
where
𝑥 is the sample mean,
𝜇 is the population mean,
𝑛 is the sample size,
𝑡 is the confidence coefficient using the 𝑡-statistic,
𝑠 is the sample standard deviation, and
𝑠
𝑡 ( ) is the margin of error.
√𝑛
𝑠 𝑁−𝑛 𝑠 𝑁−𝑛
𝑥 −𝑡( )⋅√ < 𝜇 < 𝑥 +𝑡( )⋅√
√𝑛 𝑁−1 √𝑛 𝑁−1
Example:
In a certain study, the population size is 𝑁 = 200 and the sample size is 𝑛 = 20. The
sample size is 𝑥 = 12.5 and the sample standard deviation is 𝑠 = 3.25. Determine the
length of the confidence interval of the population mean at 95% confidence level.
Since 𝑛 = 20 and 𝑁 = 200, the sample size is 10% of the population. This means that
we will use the FPC in the formula.
We are dealing with a 95% confidence level. The value of the confidence coefficient 𝑡
can be determined using the 𝑡-distribution table. To do this, let us determine first the
degrees of freedom 𝑑𝑓.
𝑑𝑓 = 𝑛 − 1
= 20 − 1
= 19
Since 𝑑𝑓 = 19 and the confidence level is 95%, we can determine the value of 𝑡 using
the 𝑡-distribution table. Note that we will have to look at the 0.025 column since this is
two-tailed.
Thus, 𝑡 = 2.093.
𝑁 = 200
𝑛 = 20
𝑥 = 12.5
𝑠 = 3.25
𝑡 = 2.093
𝑠 𝑁−𝑛 𝑠 𝑁−𝑛
𝑥 −𝑡( )⋅√ < 𝜇 < 𝑥 +𝑡( )⋅√
√𝑛 𝑁−1 √𝑛 𝑁−1
Thus, at 95% confidence level, the population mean is estimated to be between 11.05
and 13.95.
𝑝̂ 𝑞̂ 𝑝̂ 𝑞̂
𝑝̂ − 𝑧α/2 ⋅ √ < 𝑝 < 𝑝̂ + 𝑧α/2 ⋅ √
𝑛 𝑛
where
𝑝 is the population proportion,
𝑝̂ is the sample proportion with 𝑞̂ = 1 − 𝑝̂ ,
𝑧α/2 is the confidence coefficient,
𝑛 is the sample size, and
𝑝̂𝑞̂
𝑧α/2 ⋅ √ 𝑛 is the margin of error.
𝑝̂ 𝑞̂ 𝑁−𝑛 𝑝̂ 𝑞̂ 𝑁−𝑛
𝑝̂ − 𝑧α/2 ⋅ √ ⋅ √ < 𝑝 < 𝑝̂ + 𝑧α/2 ⋅ √ ⋅ √
𝑛 𝑁−1 𝑛 𝑁−1
Example:
Fifty Filipino teens were asked if they have plans to work abroad after they graduate.
It was found out that 35% of them are planning to work abroad. Determine the length
of the confidence interval of the population proportion at 99% confidence level.
Since 50 Filipino teens were asked in this study, it follows that 𝑛 = 50. The population
size is not given in the problem, but we can infer that the population refers to the total
number of Filipino teens. Thus, we can say that 𝑛 = 50 is less than 5% of the population
size, since the number of Filipino teens must be large. This means that we do not need
to use FPC in our confidence interval.
The proportion of Filipino teens who plans to work abroad is 35%. Thus, 𝑝̂ = 0.35.
Using this, we can compute for 𝑞̂ .
𝑞̂ = 1 − 𝑝̂
= 1 − 0.35
= 0.65
Since the confidence level is 99%, the value of the confidence coefficient is 𝑧α/2 = 2.58.
Using these values, we can now determine the length of the confidence interval of the
population proportion.
𝑝̂ = 0.35
𝑞̂ = 0.65
𝑛 = 50
𝑧α/2 = 2.58
𝑧α/2 ⋅ 𝜎 2
𝑛=( )
𝐸
where
𝑧α/2 is the confidence coefficient
𝜎 is the population standard deviation
𝐸 is the margin of error or the level of precision, and
𝑛 is the sample size.
Example:
Determine the minimum sample size needed to estimate the population mean of a
population with a population standard deviation of 𝜎 = 13.41. The estimated
population mean should be accurate within 3 points of the actual population mean
with a 95% confidence level.
Since the estimated population mean should be accurate within 3 points of the actual
population mean, it follows that 𝐸 = 3.
Using these values, we can now determine the minimum sample size needed to
estimate the population mean.
𝑧α/2 ⋅ 𝜎 2
𝑛=( )
𝐸
1.96 ⋅ 13.41 2
=( )
3
≈ 77
It follows that the minimum sample size needed to estimate the population mean is
𝑛 = 77.
𝑧α/2 2
𝑛 = 𝑝̂ 𝑞̂ ( )
𝐸
where
𝑧α/2 is the confidence coefficient,
𝑝̂ is the estimated proportion of an attribute,
𝑞̂ is equal to 1 − 𝑝̂ ,
𝐸 is the margin of error or the level of precision, and
𝑛 is the sample size.
In the given problem, we can determine that 𝑝̂ = 0.75 since this represents the
estimated proportion based on the previous study. Using this, we can determine the
value of 𝑞̂.
𝑞̂ = 1 − 𝑝̂
= 1 − 0.75
= 0.25
Since the estimated proportion must be within 3% of the actual proportion, it follows
that 𝐸 = 0.03. Also, since the confidence level is 95%, it follows that 𝑧α/2 = 1.96.
Using these values, we can now determine the minimum sample size needed to
estimate the population proportion.
𝑧α/2 2
𝑛 = 𝑝̂ 𝑞̂ ( )
𝐸
1.96 2
= (0.75)(0.25) ( )
0.03
≈ 800
Therefore, the minimum number of sample size needed to estimate the population
proportion is 𝑛 = 800.
Example:
A school principal wants to estimate the population mean of the scores of the students
in a national examination. She decided to use her 101 students as sample. She is able
to determine that the mean of her sample is 87.51 with a sample standard deviation
of 23.51. Estimate the population mean of the scores with a confidence level of 95%.
In the given problem, we are given with a sample size of 𝑛 = 101, a sample mean of
𝑥 = 87.51 and a sample standard deviation of 𝑠 = 23.51. The population standard
deviation is not given. Thus, the formula that we will use is as follows:
𝑠 𝑠
𝑥 −𝑡( ) < 𝜇 < 𝑥 +𝑡( )
√𝑛 √𝑛
Let us determine if we need the FPC in this problem. Note that the sample size in the
problem is 𝑛 = 101. The population size is not given. However, the context of the
problem states that the population refers to the students in the entire country since
this is a national examination. Because of this, we can safely say that the sample size
𝑛 = 101 is less than 5% of the population since the population is very large. Thus, we
do not need to use the FPC in our problem.
We can determine the confidence coefficient 𝑡 using the 𝑡-distribution table. Let us
begin by determining the degrees of freedom 𝑑𝑓.
Since we are dealing with a confidence level of 95%, we will take a look at the column
under 0.025. It follows that the confidence coefficient is 𝑡 = 1.984.
Let us determine the length of the confidence interval of the population mean.
𝑥 = 87.51
𝑡 = 1.984
𝑠 = 23.51
𝑛 = 101
𝑠 𝑠
𝑥 −𝑡( ) < 𝜇 < 𝑥 +𝑡( )
√𝑛 √𝑛
23.51 23.51
87.51 − 1.984 ( ) < 𝜇 < 87.51 + 1.984 ( )
√101 √101
87.51 − 4.6412 < 𝜇 < 87.51 + 4.6412
82.8688 < 𝜇 < 92.1512
82.87 < 𝜇 < 92.15
Therefore, at 95% confidence level, the estimated population mean is between 82.87
and 92.15.
Example:
Shannen wants to estimate the height of 6-year-old Filipino children. A previous study
has shown that the population standard deviation is 𝜎 = 8.53 cm. Determine the
minimum number of sample size needed to estimate the population mean if the
margin of error must be within 2 cm of the actual population mean and with a 95%
confidence level.
In the given problem, we know that the population standard deviation is 𝜎 = 8.53 cm.
The margin of error must be within 2 cm of the actual population mean. This means
that 𝐸 = 2. Lastly, the confidence level is 95%, which means that the confidence
coefficient is 𝑧α/2 = 1.96.
Let us determine the minimum sample size needed to estimate the population mean.
𝜎 = 8.53
𝑧α/2 = 1.96
𝐸=2
𝑧α/2 ⋅ 𝜎 2
𝑛=( )
𝐸
1.96 ⋅ 8.53 2
=( )
2
≈ 70
Thus, the minimum sample size needed to estimate the population mean is 𝑛 = 70.
𝜎 𝜎
𝑥 − 𝑧α/2 ( ) < 𝜇 < 𝑥 + 𝑧α/2 ( )
√𝑛 √𝑛
𝑠 𝑠
𝑥 −𝑡( ) < 𝜇 < 𝑥 +𝑡( )
√𝑛 √𝑛
𝑝̂ 𝑞̂ 𝑝̂ 𝑞̂
𝑝̂ − 𝑧α/2 ⋅ √ < 𝑝 < 𝑝̂ + 𝑧α/2 ⋅ √
𝑛 𝑛
○ Note that if the sample size is at least 5% of the population size, we need to
multiply the margin of error by the Finite Population Correction Factor.
𝑁−𝑛
𝐹𝑃𝐶 = √
𝑁−1
● In determining the minimum sample size needed, there are several formulas that we
can use.
○ To determine the minimum sample size needed to estimate the population
mean, we use the following formula:
𝑧α/2 ⋅ 𝜎 2
𝑛=( )
𝐸
𝑧α/2 2
𝑛 = 𝑝̂ 𝑞̂ ( )
𝐸
● In solving problems involving the sample size determination, let us remember the
following guidelines.
○ In problems involving estimating the length of the confidence interval, we
should check if we are estimating the population mean or the population
proportion. If we are estimating the population mean, we should also check if
the population standard deviation is given or not.
○ In problems involving the minimum number of sample size needed, we
should check if the given contains a population standard deviation or an
estimated proportion.
No
Yes 𝑠 𝑠
𝑥 −𝑡( ) < 𝜇 < 𝑥 +𝑡( )
√𝑛 √𝑛
𝜎 𝜎
𝑥 − 𝑧α/2 ( ) < 𝜇 < 𝑥 + 𝑧α/2 ( )
√𝑛 √𝑛
1. What term refers to the probability that the sample obtained is the representative of
the true population value?
◯ degree of variability
◯ level of precision
◯ level of confidence
◯ correlation coefficient
2. When should we use the Finite Population Correction Factor in determining the
length of a confidence interval?
◯ if the sample size is at least 5% of the population size
◯ if the sample size is at most 5% of the population size
◯ if the population standard deviation is known
◯ if the population standard deviation is unknown
3. A researcher wants to determine a sample size from a population with size 𝑁 = 1 000
with a margin of error of 5%. What is the appropriate sample size according to the
Slovin’s formula?
◯ 𝑛 = 385
◯ 𝑛 = 285
◯ 𝑛 = 384
◯ 𝑛 = 286
4. From a population of size 𝑁 = 50 000, a sample of size 𝑛 = 8 000 is drawn. What is the
value of the FPC?
◯ 𝐹𝑃𝐶 = 1.0911
◯ 𝐹𝑃𝐶 = 0.9165
◯ 𝐹𝑃𝐶 = 0.8400
◯ 𝐹𝑃𝐶 = 1.1905
6. In a certain study, the estimated proportion is 𝑝̂ = 0.3. What is the minimum number
of sample size needed to estimate the population proportion if the estimated
proportion must be within 2% of the actual proportion and with a confidence level of
95%?
◯ 𝑛 = 1 183
◯ 𝑛 = 3 495
◯ 𝑛 = 2 017
◯ 𝑛 = 683
7. Macy plans to conduct a study on the living conditions of prisoners in a local jail. She
plans to interview some prisoners. From a list of 200 prisoners, she plans to select
every 15th person from the list. What kind of sampling procedure is being shown in
this situation?
◯ Simple random sampling
◯ Stratified random sampling
◯ Systematic random sampling
◯ Quota sampling
9. A candy company wants to test if their new candy will be liked by Filipino children.
They selected 50 children as participants of the study. Based from their test, 37 of
them likes the new candy. What is the estimated proportion of Filipino children who
likes their new candy at 99% confidence level?
◯ between 58.00% and 90.00%
◯ between 61.84% and 86.16%
◯ between 13.84% and 38.16%
◯ between 10.00% and 42.00%
10. Mina wants to estimate the mean of the standby time of a certain brand of phone.
She wants her estimate to be within 0.2 hours from the actual population mean at
95% confidence level. Based from the previous study, the population standard
deviation is 𝜎 = 0.81 hours. How many phones should she test for her study?
◯ 109 phones
◯ 63 phones
◯ 89 phones
◯ 45 phones
Tips
On Selecting a Sampling Procedure
● In determining the sampling procedure for your study, make sure to consider the
variables of your study. The randomness in selecting the participants of your study is
necessary so that bias in selecting participants would be avoided.
● As much as possible, we try to avoid selecting participants just because it is
convenient to the researcher, but there are instances where it cannot be avoided,
especially if we are restrained by time or money.
References
Bhat, Adi. “Quota Sampling: Definition, Steps, Examples and Advantage,” Question Pro.
Accessed from https://www.questionpro.com/blog/quota-sampling/, 28 December
2019.
Glen, Stephanie. “Slovin’s Formula: What is it and When do I use it?” Statistics How To.
Accessed from https://www.statisticshowto.datasciencecentral.com/how-to-use-
slovins-formula/, 28 December 2019.