HOW TO DETERMINE THE SAMPLE SIZE

USING THE SAMPLE WEB

The recommended sample size for a given population size, level of confidence, and
margin of error appears in the body of the table.
Change these values to select different levels of confidence.
Change these values to select different maximum margins of error.
Change these values to select different (e.g., more precise)
population sizes.

This the Sample Web. (It is in excel format,”SampleSize-web”). You are required to
make use of the template for sample size determination given necessary conditions,
level of confidence, and |margin if error (degree of accuracy)

1. For example, the recommended sample size for a population of 50, a confidence
level of 99%, and a margin of error (degree of accuracy) of 0.035 or 3.5% is 48.

2. Next example, given a population of 250, with a confidence level of 95%, and a
margin of error of 1%, the sample size is 244.
3. What if the given confidence level is 95%, having a population of 850? What is the
sample size if the margin of error is 5%? 3.5%? 2.5%? 1%?

Based on the picture, under the column, Population Size, 850 is nowhere to find and is
actually between 800 and 900. As per said, this can be edited to get the precise number
of sample size. Therefore, change the value of 800 to 850, and directly, sample size will
be provided on the different margin of errors.

There, the sample sizes for a population of 850 with margins of error of 5%, 3.5%,
2.5%,1%, were 265, 408, 547, and 782, respectively.

4. For the next example, lets us try to manipulate the sample web when the confidence
level is neither 95% nor 99%. Lets us try other levels. Using a number of population of
1,200, let’s find the sample size of the different margins of error with a confidence level
of 89%.
Solution: You just have to change the value of the confidence level which is 95% into
89%.

Therefore, the sample sizes having a confidence level of 89% with margins of error of
5%, 3.5%, 2.5%,1%, are 211,364,552, and 1010, respectively.

5. Now, let us try to manipulate the margins of error as well. Given a population size of
7,546, with a confidence level of 90%, find the sample size using the margins of error
of 4%, 6%, 9%, and 10%.

Solution: We need to change the population size into 7,546; confidence level of 90%;
and margins of error of 4%, 6%, 9%, and 10%.

Given the population of 7,546, the sample sizes having a confidence level of 90% with
margins of error of 4%, 6%, 9%, and 10%, are 400, 183, 83, and 67, respectively.
DETERMINING SAMPLES
A. Determining samples from strata
(Stratified Sampling)
Consider the data below. Let’s determine the number of samples to be selected
from each group(strand). Use 99% level of confidence and 0.01 margin of error.
(Sample web is required to determine the sample size)

STRAND N Percentage Samples (n)

ABM 228
STEM 320
HUMSS 140
GAS 312
Total 1000

Steps:
a. Complete the table by identifying the percentage of each strand.
STRAND Population (N) Percentage Samples (n)
ABM 228 23%
STEM 320 32%
HUMSS 140 14%
GAS 312 31%
Total 1000 100%

228 140
P ABM = =0 . 23=23 % PHUMSS = =0 .14=14 %
1000 1000
320 312
PSTEM = =0 . 32=32 % PGAS = =0 .31=31 %
1000 1000
*The total percentage is 100%.
 b. Determine the total number of samples using the sample web. There are 1000
total population. Type 1000 on the violet column. The computation is
automatically done. The needed number of samples is the intersection of 1000
and 0.01 margin of error under 99% confidence. 943 is the required samples.

This means that we need to distribute 943 among four strands. In doing this, we
need to specify the percentage per strand.

c. Determine the number of samples (n) from each strand multiplying the
percentage of each strand/track to the total samples, 943.

Strand Population (N) Percentage Samples (n)

ABM 228 23% 217
STEM 320 32% 302
HUMSS 140 14% 132
GAS 312 31% 292
Total 1000 100% 943

n ABM =0 .23∗943=216 .89≈217

n STEM =0 .32∗943=301 . 76≈302
n HUMSS=0. 14∗943=132. 02≈132
nGAS =0. 31∗943=292 .33≈292
Notice that the group with the largest number of students will have the largest
number of samples. 302 samples will come from STEM, while 292, 217, and 127
samples will come from GAS, ABM, and HUMSS strands.

Note: The table above is limited only to the number of samples to be selected from
each strand without recognizing other subgroups on each group/strand. Possible
subgroups are grade level and sections per grade level per strand.
B. Determining samples from cluster
(Cluster Sampling)
In the previous example, samples are distributed to strands only.
Strand Population (N) Percentage Samples (n)
ABM 228 23% 217
STEM 320 32% 302
HUMSS 140 14% 132
GAS 312 31% 292
Total 1000 100% 943

What if we desire to determine the number of samples per grade level?

Consider the data below.
Grade No. of No. of
STRAND Percentage Sample (n)
Level Sections Students
ABM 11 3 116
12 3 112
Total 6 228
STEM 11 4 160
12 4 160
Total 8 320
HUMSS 11 2 70
12 2 70
Total 4 140
GAS 11 4 156
12 4 156
Total 8 312
TOTAL 26 1000

Reminder:
Our intention is to distribute the samples per strand to each grade level by
recognizing the table under stratified sampling.
STEPS.
1. Determine the percentage of the grade level per strand. In getting the
percentage for each grade level, recognize the total number of students in
the strand.
116 160
P ABM( Grade 11)= =0 . 51=51 % PSTEM (Grade 11)= =0 . 50=50 %
228 320
112 160
P ABM( Grade 12)= =0 . 49=49 % PSTEM (Grade 12)= =0 .50=50 %
228 320
 70 156
PHUMSS( Grade11 )= =0 .50=50 % PGAS( Grade11 )= =0 . 50=50 %
140 312
70 156
PHUMSS( Grade12 )= =0.50=50 % PGAS( Grade12 )= =0 .50=50 %
140 312
To determine the percentage of each group, recognize the total number of
students which is 1000.

228 140
P ABM = =0 . 23=23 % PHUMSS = =0 .14=14 %
1000 1000
320 312
PSTEM = =0 . 32=32 % PGAS = =0 .31=31 %
1000 1000

Grade No. of No. of

STRAND Percentage Sample (n)
Level Sections Students
ABM 11 3 116 0.51
12 3 112 0.49
Total 6 228 0.23
STEM 11 4 160 0.50
12 4 160 0.50
Total 8 320 0.32
HUMSS 11 2 70 0.50
12 2 70 0.50
Total 4 140 0.14
GAS 11 4 156 0.50
12 4 156 0.50
Total 8 312 0.31
TOTAL 26 1000 100%
2. After completing the percentages, we are ready to determine the samples per
grade level per strand. For example, in ABM we need 217 samples to be
identified from two grade levels. To get the samples from Grade 11 ABM,
multiply the grade 11 percentage to 217 which is rounded off to 110. Grade
12 ABM sample size is 107. Others will follow. In symbols,
n ABM (Grade 11)=0 .51 x 217=110.404≈110
n ABM (Grade 12 )=0 .49 x 217=106 .596≈107
n STEM (Grade 11)=0 .50 x 302=151
n STEM (Grade 12 )=0 . 50 x 302=151
n HUMSS(Grade 11)=0 .50 x132=66
n HUMSS(Grade 12)=0 . 50 x 132=66
nGAS (Grade 11)=0 .50 x292=146
nGAS (Grade 12)=0 . 50 x 292=146

Grade No. of No. of

STRAND Percentage Sample (n)
Level Sections Students
ABM 11 3 116 0.51 110
12 3 112 0.49 107
Total 6 228 0.23 217
STEM 11 4 160 0.50 151
12 4 160 0.50 151
Total 8 320 0.32 302
HUMSS 11 2 70 0.50 66
12 2 70 0.50 66
Total 4 140 0.14 132
GAS 11 4 156 0.50 146
12 4 156 0.50 146
Total 8 312 0.31 292
TOTAL 26 1000 100% 943

This table provides the scheme in determining the number of samples from
grade levels grouped per strand.
C. Determining samples from cluster
(Multi-Stage Sampling)
In the previous example, samples are limited to strands and then grouped to
grade level. Let’s review them

Table 1. Samples distributed per strand. (STRATIFIED)

Strand Population (N) Percentage Samples (n)

ABM 228 23% 217
STEM 320 32% 302
HUMSS 140 14% 132
GAS 312 31% 292
Total 1000 100% 943

Table 2. Samples from each strand distributed to grade levels. (CLUSTER)

Grade No. of No. of

STRAND Percentage Sample (n)
Level Sections Students
ABM 11 3 116 0.51 110
12 3 112 0.49 107
Total 6 228 0.23 217
STEM 11 4 160 0.50 151
12 4 160 0.50 151
Total 8 320 0.32 302
HUMSS 11 2 70 0.50 66
12 2 70 0.50 66
Total 4 140 0.14 132
GAS 11 4 156 0.50 146
12 4 156 0.50 146
Total 8 312 0.31 292
TOTAL 26 1000 100% 943

What if we desire to determine the number of samples from each section?

This is multi-stage sampling. To do this, we will determine the needed number of

samples from each section per grade level per strand. We are distributing the
sample size to each section from a specific sample size from the grade level per
strand.
STRAN No. of
Grade Level Sections Percentage Sample (n)
D Students
A 39 0.34 0.34 x 110 =37.4 37
B 39 0.34 0.34 x 110 =37.4 37
11
C 38 0.33 0.33 x 110 =36.3 36
Total 116 0.51 110
ABM A 36 0.32 0.32x107= 34.24 35
B 38 0.33 0.33x107=35.31 36
12
C 38 0.33 0.33x107=35.31 36
Total 112 0.49 107
Total 228 0.23 217
A 37 0.23 0.23x151=34.92 35
B 42 0.26 0.26x151=39.64 39
11 C 41 0.26 0.26x151=38.69 39
D 40 0.25 0.25x151=37.75 38
Total 160 0.50 151
STEM A 40 0.25 0.25x151=37.75 38
B 40 0.25 0.25x151=37.75 38
12 C 40 0.25 0.25x151=37.75 38
D 40 0.25 0.25x151=37.75 37
Total 160 0.50 151
Total 320 0.32 302
A 34 0.49 0.49x66 = 32.06 32
11 B 36 0.51 0.51x66 = 33.94 34
Total 70 0.50 66
HUMSS A 35 0.50 0.50x66 = 33 33
12 B 35 0.50 0.50x66 = 33 33
Total 70 0.50 66
Total 140 0.14 132
A 39 0.25 0.25x146 = 36.5 36
B 39 0.25 0.25x146 = 36.5 36
11 C 39 0.25 0.25x146 = 36.5 37
D 39 0.25 0.25x146 = 36.5 37
total 156 0.50 146
GAS A 39 0.25 0.25x146 = 36.5 36
B 39 0.25 0.25x146 = 36.5 36
12 C 39 0.25 0.25x146 = 36.5 37
D 39 0.25 0.25x146 = 36.5 37
total 156 0.50 146
Total 312 0.31 292
Grand Total 1000 100% 943
After determining the needed sample size,
getting/identifying the samples will be the next step
to do. To perform this, SAMPLING TECHNIQUES will
be applied. Some of them are lottery method, table of
random numbers, to name a few.

Imagine this scenario, you own a grocery store.

50 boxes containing 100 cans of sardines are
delivered. You’d like to check if all boxes contain
sardines of common expiration date as suggested.
You have a total of 5000 cans of sardines. In a
normal setting, you can’t check all cans so you prefer
to get samples. You will get samples from each box,
knowing the chance that each box may contain
different expiration. And this samples will represent
the entire number. If you found out expired canned
sardines, then you will refuse to purchase and sell the
product. Sampling is essential to quality control.