SAMPLING

Definition
Sampling may be defined as measuring a small portion of something md then making a general
statement about the whole thing (Bradfield and Moredock. p.38) For instance, one is buying
mangoes. He picks up one mango from a crate of mangoes and tastes it. If the mango is sweet,
then he concludes and says that the mangoes in the crate are sweet although he has tasted only
one mango.
Why We Need Sampling (Purposes and Advantages of Sampling)
There are some advantages of sampling. some of which are the following:
1. Sampling makes possible the study of a large, heterogeneous population. The universe or
population to be studied may be too large or unlimited that it is almost impossible to
reach all of them. Suppose the age distribution of the entire population of the country is
to be studied. It is surely very difficult if not entirely impossible to reach every person of
the whole country. Thus, sampling makes possible this kind of study because in sampling
only a small portion of the population may be involved in the study, enabling the
researcher to reach all through this small portion of the population.
2. Sampling is for economy. Research without sampling may be too costly. Take for
instance that a study involves the whole population of a town of 70,000. To make
questionnaire copies for 70,000 is surely an expensive affair. The expenses are even
greater if interviewers are employed to interview every person of that group numbering to
70,000. However, sampling re duces the study population to a reasonable size that
expenses are greatly reduced.
3. Sampling is for speed. Research without sampling might be too time consuming. To
reach the 70,000 persons for an interview would surely entail a long time unless many
interviewers are employed. If a research takes a long time to finish, there may be many
intervening factors that deter the researcher from finishing his research.
4. Sampling is for accuracy. If it takes too long a time to cover the whole study population
there may be inaccuracy. By the time the last person is interviewed, the data gathered
from the first interviewees may be obsolete already so that the conclusions are no longer
accurate. The research must be finished within a reasonable period of time so that the
data are true, valid, and reliable.
5. Sampling saves the sources of data from being all consumed. The act of gathering data
may consume all the sources of information without sampling. Suppose we want to find
out whether bread baked in a certain bakery is delicious. Without sampling, we have to
consume all the bread baked in the bakery before making any conclusion. In such a case,
there is no more bread to apply the conclusion to.
Sampling Concepts and Terminology (Babbie, pp. 141-144)
All definitions here are from Babbie although explanations are added for local adaptation.
Element. An element is that unit about which information is collected and which
provides the basis of analysis. They are the members of the population. These are certain
types of people, families, social clubs, and the like. Elements and units of analysis are
often same, but the former refers to sample selection while the latter refers to data
analysis. For instance. the families in a slum area are the subject of an investigation.
While selecting the families to be included in the sample, they are called elements but
when data have already been gathered and being analyzed, the families are called units of
data analysis.
Population. A population is the theoretically specific aggregation of the elements. This is
also called universe. Suppose a study is to be conducted to determine the profile of the
college students of a university. The term population includes all the college students
whether full time, part time, professional, working, male, or female students.
Study population. A study population is that aggregation of elements from which the
sample is actually selected. In the study of college students, only the full time, male and
female students may be included, all other types of students excluded. This is the study
population which is often defined in the scope and delimitation of a thesis report.
Sample. These are the elements (people) who are actually selected to participate or to be
the subjected in the study.
Sampling unit. A sampling unit is that element or set of elements considered for
selection in some stage of sampling. In a simple, single-stage sampling, the sampling
units are the same as elements. However, in a more complex sampling, different levels of
sampling units may be employed. For example. you may select a sample of census blocks
in a city, then select a sample of adults, of which only the last of these are the elements.
The first sampling unit is called primary sampling unit, the census blocks; the second unit
is called the secondary sampling unit. the household; and the third unit is called the final
sampling unit, the adults.
Sampling frame. A sampling frame is the actual list of sampling units from which
sample, or some stage of the sample, is selected. If a sample of students is selected from a
student roster, the roster is the sampling frame. In a single-stage sampling design, the
sampling frame is the list of the elements composing the study population.
Observation unit. An observation unit, or unit of data collection is an element or
aggregation of elements from which information is collected. The unit of analysis
observation unit are often the same —the individual person — but not always the same.
Thus, the researcher may interview the heads of households (the observation units) to
collect information about all members of the households (the units of analysis). In other
words. the observation units are the actual respondents to a study. The observation units
or respondents are the people who arc interviewed or who are requested to accomplish
questionnaires for data collection.
Variable. A variable is a set of exclusive attributes. Examples are sex, age, employment
status, and so forth. The elements of a given population may be described in terms of
their individual attributes on a given variable. Thus, a researcher may describe the age
distribution of a population by examining the relative frequency of their ages. A variable
must possess variation. If all elements in the population possess the same attribute, that
attribute is constant in the population, rather than a part of a variable.
Parameter. A parameter is summary description of a given variable in a population.
The mean income of all the families of a city is a parameter. The mean age of all the
people in a city is also a parameter. The age distribution of all the people is a parameter.
Statistics. A statistics is the summary description of a given variable in a sample. Thus,
the mean income computed from a sample the age distribution of that sample are
statistics. Sample statistics are used to make estimates of population parameters.
Sampling error. Sampling error is the degree of error of a sample statistic when
compared with the population parameter. For instance, if the population mean is 80 and
the sample mean is 40, the sampling error is 40. The acceptable maximum sampling error
is 5%
Representative sampling. A sample will be representative of the population from which
it is selected if the aggregate characteristics of the sample closely approximate those
same aggregate characteristics in the population. Put in another way, a sample will be
representative of the population from which it is selected if all members of the population
have an equal chance of being selected in the sample. This means that characteristics
(differentiated groups or classes in the population) must be proportionally represented. If
there are 50% women in the study population is 100, 60 of whom are women and 40 are
men. If the sample is 20% of the population, then there should be 12 women are 8 men in
the sample.
Confidence level. Confidence level is the degree of confidence that a sample statistic will
accurately fall within a certain or specified interval from the population parameter. The
interval, called the confidence interval, is the range within which the sample statistic
supposes the sample mean is 75 and the confidence interval is plus or minus 8, and the
level of confidence is 95%. Then we say that we are 95% confident that the population
mean falls within plus or minus 8 to or from 75. Hence, the confidence interval or range
is from 67 to 83. (75-8=67 and 75+8=83)
Sampling stratum. (an addition) A sampling stratum is that group, selection, or category
of elements from which selection is made in some stage of sampling. In a single stage
sampling, the sampling stratum is the whole population. However, in a more complex
sampling, different levels of sampling strata may be employed. For example, you may
select a sample of census blocks in a city, then select a sample of households from the
selected blocs, and finally select a sample of adults from the selected households. The
city is the primary sampling stratum, the census blocks are the secondary sampling
stratum, and the households are the final sampling stratum.
Principles of Sampling (Bradfield and Moredock, pp. 39-40)
1. Appraisals that involve sampling are estimates and predictions only.
2. Estimates based on sampling are least accurate when the sample is a small proportion of
the whole and when the sample is not representative. Conversely, estimations based on
proportionately large samples and on representative samples are most accurate.
(Sample means the portion of a population taken as the subject of a study. A sample is
representative if it is taken from all sections or groups of a population in equal proportions.)
3. Sampling may be categorical or temporal. Sampling is categorical if the sample is taken
proportionally from categories or groups. Sampling is temporal when the sample is in
terms of time, as for example the pulse of a patient is taken every ten minutes.
Disadvantages of Sampling (Defective Sampling)
1. If sampling is biased, or not representative, or too small, the conclusion may not be valid
and reliable.
Here is an example of a representative sample. Suppose the teaching of science is to be studied in
a certain high school with the students from first year to fourth as the respondents. The
population or universe consists of all the high school students. A portion of 20% are to be taken
as respondents. This constitutes the sample. Suppose the students are divided into sections
according to ability. The diagram below illustrates how a representative sample is taken. Take
the first year students who are sectioned into three according to ability.
Number 20%
Male 20 4
Section 1
Female 30 6
Male 20 4
Section 2
Female 25 5
Male 18 4 (rounded)
Section 3
Female 27 5 (rounded)
Total 140 28

As far as the first year students are concerned, 140 is the population and 28 is the sample which
is representative. The same should be done with the second, third, and fourth years.
If the students are lumped up together (combined) and then the 20% will be taken from the
whole, the sampling is biased and not representative because more might be taken from one
section or there may be more males than females, etc.
2. In research, the respondents to a study (the sample) must have a common characteristic
which is the basis of the study. If some of the sample do not have this common
characteristic, the conclusions become faulty. In the example given in No.1, the common
characteristic of the sample or population is that all of them are high school students. If
some college students enter the sample, the conclusions may become erroneous.
 3. If the population is very large and there are many sections and subsection, the sampling
procedure becomes very complicated. It would be difficult to assign numbers to the
elements (members) of the population or it would also be very difficult to write their
names alphabetically on pieces of paper to be placed in the raffle box.
4. If the researcher does not possess the necessary skill and technical knowhow in sampling
procedure, the sampling may become biased and unrepresentative.
General Types of Sampling
There are two general types of sampling (1) probability sampling and (2) non-probability
sampling.
Probability sampling. In probability sampling, the sample is a proportion (a certain
percent) of the population and such sample is selected from the population by means of
some systematic way in which every element of the population has a chance of being
included in the sample.
Non-probability sampling. In non-probability sampling, the sample is not a proportion
of the population and there is no system in selecting the sample. The selection depends
upon the situation.
Types of Non-probability Sampling
There are three types of non-probability sampling: (1) accidental sampling, (2) quota sampling,
and (3) convenience sampling.
Accidental (Incidental) sampling. In this type of sampling, there is no system of
selection but only those whom the researcher or interviewer meet by chance are included
in the sample. For instance, the interviewer stands in a street corner and he interviews
everyone who passes by. The problem with this type of sampling is its lack of
representativeness. The sample might be a biased one. If the place where the interviewer
stands is a business section, most of the people who pass by are business and rich people,
hence, the respondents are only well-to-do people. If the place where the interviewer
stands is a slum are, most of the people who pass by and interviewed are poor people. In
research, every section of the population being studied must be proportionally
represented in the sample.
Accidental sampling is resorted to only when there is no other alternative.
Quota sampling. In this type of sampling, specified numbers of persons of certain types
are included in the sample. Suppose the reactions of the people towards a certain issue is
to be determined and a cross-section of the population must be involved. It might be
decided that ten doctors, eight lawyers, fifteen businessmen, twenty vendors and twenty-
five farmers should constitute the sample.
The advantage of quota sampling over accidental sampling is that many sectors of the
population are represented. However, its representativeness is doubtful because there is
no proportional representation and there are no guidelines in the selection of the
 respondents. Anyone of the groups selected to participate will do. Any doctor or an
lawyer who like to participate will be included in the sample. Another danger is that the
perception of the minority may become typical in the findings. If the minority of the
sample, for example, are professionals who are in the minority of the population, their
perceptions may become predominant over those of the non-professionals who are in the
majority of the population.
Again, quota sampling may be used only when any of the more desirable types of
sampling will not do.
Convenience sampling. Convenience sampling is a process of picking out people in the
most convenient and fastest way to immediately get their reactions to a certain hot and
controversial issue. If there is a hot and controversial issue, say the raising of the price of
gasoline, the most convenient and fastest way of reaching people is by the telephone to be
able to interview them at once. So the interviewer contacts people who have telephones
and asks them what their reactions are to the issue at hand.
This type of sampling is certainly biased and not representative considering that the
people who have telephones are a class by themselves and so their views cannot be
considered as views of the people. It should be notes that even in ordinary data gathering
in research, interviews can also be conducted through the telephone provided that those
interviewed by telephone are only among those who are properly selected too be included
in the sample. It only happens that these people have telephones.
Types of Probability Sampling (Techniques of Sampling)
1. Pure random sampling. This type of sampling is one in which everyone in the
population of the inquiry has an equal chance of being selected to be included in the
sample. This is also called the lottery or raffle type of sampling. This may be used if the
population has no differentiated levels, sections, or classes.
Usually, this is done this way. A certain percent of the population is to be selected. This is
decided upon beforehand. Suppose there are 200 persons in the population and 20 percent is the
proportion to be chosen. Hence, 20 percent of 200 equals 40. The names of all the 200 persons
are written each on a piece of paper or if they are given numbers only their numbers are written.
The pieces of paper are then placed in a container convenient for the purpose, say a box. The
pieces of paper which are usually rolled for secrecy are mixed thoroughly and the 40 pieces are
draw by chance. Those whose names or numbers are drawn are included in the sample.
The main advantage of this technique of sampling is that it is easy to understand, and it is easy to
apply too. The main disadvantage is that it is easy to apply too. The main disadvantage is that it
is hard to use with too large a population because of the difficulty encountered in writing the
names of the persons involved. However, this problem may be offset by the use of a table of
random numbers. The use of a table of random numbers are explained in good statistical books.
(Peatman, pp.181, 204-207)
Pure random sampling is also called unrestricted random sampling, which means that every
individual in the population has an equal chance of being chosen to be included in the sample.
2. Systematic sampling. This is a technique of sampling in which every nth name in a list
may be selected to be included in a sample. This is used when the subjects or respondents
in the study are arrayed or arranged in some systematic or logical manner such as
alphabetical arrangement, residential or house arrays, geographical placements from
north to south, etc.
The procedure follows. Suppose 20% of the population is the sample size. If 100% is divided by
20%, the answer is 5. Hence, every fifth name in the list or array is selected. However, there
must be a random start. The researcher may close his eyes and run his finger points to at this
particular time is the random start number. Suppose this number is 13. This is the first selection.
The succeeding numbers to be selected are found by adding 5 to its predecessor. So the next
number to be selected is 18(13+5), the next is 23(18+5), the next number is 28(23+5), etc.
The main advantage of this technique of sampling is that it is more convenient, faster, and more
economical than the pure random sampling. The main disadvantage is that the sample becomes
biased if the persons in the list belong to a class by themselves whereas the investigation requires
that all sectors of the population are to be involved. Suppose the inquiry is about agrarian reform
and the investigator selected his respondents from a list of landowners or land tax payers
obtained from the Land Tax Office of a town. Naturally, these people are against land reform,
unless the researcher wants to find out how many of the land tax payers are in favor of agrarian
reform.
The systematic random sampling is a restricted random sampling because there are certain
restrictions imposed upon it.
3. Stratified random sampling. Stratified random sampling is the process of selecting
randomly, samples from the different strata of the population used in the study. Thus, this
is used when the population of the inquiry has class stratifications or groupings either
horizontally or vertically. Examples of horizontal stratifications are sections in the same
school grade or year such as Section 1, Section 1, etc., sex such as males and females,
college course such as AB, BSC, BSN, BEED, BSED, etc. Examples of vertical
stratification are age such as 7, 8, 9, 10 years old, grades in school such as Grade 1,
Grade 2, Grade 3, etc., economic status such as poor, average, rich, etc.,
The procedure of stratified random sampling follows. Determine a common stratum or class to
which all the elements of the population belong. Then divide or group the elements of the
population belong. Then divide or group the elements of the population according to the
characteristics inherent in the whole class or stratum that make the elements of the population
different from one another. For example, the respondents to a study are college students. The
common class or stratum that all of them belong to is college or college studies. The
characteristics that make them different from one another are the courses they are taking, such as
BEED, BSED, BSN, etc.
Other characteristics are their curricular years, such as first, second, third, or fourth year.
Another factor is sex; still another is age. After stratification has been completed, apply either the
pure random or systematic random sampling in the actual selection of the samples in every
section or stratum and the selection must be proportional, that is, the same percent is used in
every section or class irrespective of the number in each section or class.
This is another example of stratified random sampling and its advantage is that it contributes
much to the representativeness of the sample. It is also easy to apply. Here is an illustration of
the technique. Suppose the students of a college are respondents in a study. The students are
stratified according to the courses they are taking, their sex, and the curricular years they are in.
The sample of 20% is taken from every stratum based on course, sex and curricular year. See the
table immediately following. For instance, there are 50 male students in the first year taking
education. The sample is 20% of 50 is 10. There are 380 female students in the second year
taking commerce. The sample is 20% of 380 is 76.
Curricular Years
Total
I II III IV
Course Sex
20 20
Fr 20% Fr 20% Fr Fr Fr 20%
% %
M 50 10 40 8 32 6 28 6 150 30
Education
F 75 15 60 12 53 11 50 10 238 48
Commerc M 300 60 270 54 250 50 240 48 1060 212
e F 400 80 380 76 350 70 330 66 1460 292
M 25 5 24 5 22 4 20 4 91 18
Nursing
F 300 60 260 52 255 51 250 50 1065 213
115 103
Total 230 207 962 192 918 184 4064 813
0 4

Note: Fr = Frequency, number of students (population), total = 4064, M = Male students, F =

Female students, 20% = Sample proportion, all frequencies under 20% are included in the
sample, total = 813. The selection in every group or stratum, either the pure random sampling or
the systematic random sampling may be used.
4. Purposive sampling. Purposive sampling is determining the target population, those to
be involved in the study. The respondents are chosen on the basis of their knowledge of
the information desire. If research is to be conducted on the history of a place, the old
people of the place must be consulted. If methods and techniques of teaching are the
subjects of an inquiry, teachers are the ones contacted. If the topic of investigation is the
production of rice, the respondents must be rice farmers and agriculturists. However, the
actual sampling or systematic random sampling. If the population is composed of
differentiated groups, stratified random sampling must be utilized.
5. Cluster sampling (Multistage cluster sampling). Cluster sampling or multistage cluster
sampling or simply multistage sampling is used when the population is so big, or the
geographical area of the research is so large. The general procedure is to divide the area
 or population into clusters or blocks and then within the final cluster apply any of the
different methods of selecting a sample. Suppose a study is to be conducted to determine
the standard of instruction in a certain region of the country in elementary schools. Take
for granted that there are nine provinces in the region and the proportion of the
population to be sample is 20%. The steps are:
a. Make a listing of the sampling units, the primary sampling units (the first
clusters to be sampled), the secondary sampling units (the second clusters to
be sampled within the primary sampling unit), etc. Suppose the provinces are
the primary sampling units, the towns are the secondary sampling units, and
the barangays as the final sampling units. These are called natural clusters.
b. Since the sample is 20%, 20% of 9 provinces equals 1.8 or two provinces.
Select these two provinces by either pure random or systematic random
sampling.
c. Within each of these two provinces, select 20% of the towns either by pure
random or systematic random sampling.
d. Within each town selected, choose 20% of the barangays or barrios. Since
there is only one elementary school in one barrio or barangay, this is the final
sampling unit or cluster. The respondents in the study are to be selected from
these final clusters. The respondents may be stratified into teachers,
administrators, parents and pupils. The respondents have to be taken from
these stratified groups by either pure random or systematic random sampling.
The advantage of the multistage cluster sampling is its efficiency. Its main disadvantage is its
reduced accuracy or representativeness, on account of the fact that in every stage there is a
sampling error. (For a more detailed and clearer explanation of the multistage cluster sampling,
see Babbie, pp.162-169)
Guidelines for Determining Adequate Sampling
The size of a sample is usually determined before the conduct of any study. There are no fixed
rules in determining the size of a sample needed. However, there are broad guidelines that should
be observed in determining the size of a sample. Among these guidelines are:
1. When the universe or population is more or less homogeneous and only the typical,
normal, or average is desired to be known, a smaller sample is enough. However, if
differences are desired to be known, a larger sample is needed.
2. When the population is more or less heterogeneous and only the typical, normal, or
average is desired to be known, a larger sample is needed. However, if only their
differences are desired to be known, a smaller sample is sufficient.
3. The size of a sample varies inversely as the size of the population. A larger proportion is
required of a smaller population and a smaller proportion may do for a bigger population.
For a population of five thousand, a sample of 10 percent may do but for a population of
five hundred, a proportion of 30 percent may be required.
4. For a greater accuracy and reliability of results, a greater sample is desirable.
 5. In biological and chemical experiments such as testing the effects of drugs and other
substances, the use of a few persons is more desirable to determine the reactions of
humans to such drugs and other substances being tested.
6. When the subjects (sample) are likely to be destroyed during the experiment, it is more
feasible to use non-humans such as animals especially rats.
Steps in Computing the Size of a Sample.
The steps in computing the size of a sample are as follows.
1. Determine the size of the study population. This is easily determined from the scope and
delimitation of the study.
2. Decide on the margin of error. As much as possible the margin of error should not be
higher than 5%. Probably, 3% is an ideal one.
3. Use the formula n=N/(1+Ne2) (Pagoso, et al., p.46)
In which n = the size of the sample, N = the size of the population, e = the margin of error
4. If the sampling is multistage or if the population is stratified, compute the sample
proportion (percent) by dividing the result in Step No.3 by the population.
5. Multiply the number of sampling units in each final sampling stratum by the rate
(percent) to find the sample from each final sampling stratum.
6. Add the samples from all the final sampling strata to find the total sample.
Example. Suppose an investigation of the teaching of science in the high schools of a region is to
be conducted, in which the science teachers are to be made respondents. There are 2243 teachers
handling biological sciences. 1406 handling chemical sciences, and 992 teachers handling
physical sciences, a total of 4641 teachers.
The sampling procedure follows:
Step 1. The population is 4,641
Step 2. The margin of error to be used is 3%
Step 3. Using the formula
n = 4641/ 1+4641(.03)^2
n = 4641/ 1+4641(0.0009)
n = 4641/ 1+4.1769
n = 4641/ 5.1769
n = 896.4825 (this may be rounded to 900 for convenience) This is the sample
Step 4. The teachers are grouped into three categories according to the branch of science they are
handling, so we use stratified sampling.
Sample proportion (%) = n/N = 900/4641 = 0.1939
 0.1939 may be rounded off to 0.20 or 20% for convenience of computation
Step 5 and 6
Teachers handling Number % Sample
Biological Sciences 2243 20 449
Chemical Sciences 1406 20 281
Physical Sciences 992 20 198
Total 4641 928

The total sample is 928