Sampling Methods / Techniques: Probability vs Non-Probability Sampling

Sampling is one of the most important factors which determine the accuracy of your survey result. If the population under study is small, then there is a possibility to collect the data for each item or entity. But this is rarely the situation in survey research as most of the time the universe of population under study is tool large which necessitates the need to select a sample for the study. Sampling helps researchers to save time, energy, and money that require studying the whole population. It helps them to get inferences applicable to whole population with minimum resources.

Definition: Sampling is the process of selecting a representative group from the population under study with the intention of finding out a conclusion which can be applied to total population from which they were chosen. In short, it is the process of obtaining information about an entire population by examining only a portion of it. The method through which a sample is selected is known as ‘Sampling methods’. Let’s start with the understanding of basic terms in this context: 

Population: The entire group of people or items that the researcher wants to study. It is denoted by ‘N’ and alternatively known as 'Universe' also.

Sample: Samples are drawn from populations. The portion of the population, the researcher select for the study. It is denoted by ‘n’.

Sampling: The process of selecting a group of individuals from a population in order to study them and characterize the population as a whole.

Need for Sampling: The need for sampling is felt due to the following reasons:-

i) Economy: Sampling saves time, efforts and money
ii) Timeliness: It helps to produce results relatively faster than a census study
iii) Large size of many populations: Sampling can be the only choice when population contains infinite numbers.
iv) Inaccessibility of population: Sampling helps in the case of geographically scattered diverse population
v) Accuracy: Sampling may help to get more accurate measurements if conducted by trained and experienced investigators.

Characteristic of a Good Sample: Validity of research results depends highly upon the quality of the sample drawn. If the sample is biased and does not truly represent the population, then the results cannot be trusted or generalized. There are certain factors that can affect the inferences drawn from a sample. For example: (a) the larger the sample, the more is the chance of accuracy (b) the higher the variations in population, the greater is the chance of uncertainty in outcome, however on the contrary, the higher the consistency in population, the greater is the chance for quality of outcome, and (c) the higher the variations in population, the larger should be the size of the sample.

A sample should essentially possess the following characteristics:
i) True representative of the Population: A sample should homogeneously represent the population it is taken from.
ii) Goal-oriented: A sample should be oriented to research objectives and fitted to the survey conditions
iii) Accuracy: It should be free from error and bias
iv) Economical & Practical: The sample should be such that the objectives of research can be obtained with minimum cost and effort. At the same time, it should be simple and practical to be followed easily.
v) Proportional: Sample should be adequate and proportional in size and reliable

Types of Sampling: Sampling techniques/methods can be categorized into the following two broad types:

PROBABILITY SAMPLING: Probability sampling utilizes Random Sampling Techniques to create a sample. This group of sampling methods gives all the members of a population equal chance of being selected. This means that the selection of sample is independent of the person making the study and the sampling operation is controlled so objectively that all items will be chosen strictly at random. Probability sampling is more complex, time-consuming and generally more costly than non-probability sampling. Probability sampling can be further grouped into following types:

a)      Simple Random Sampling: Simple Random Sampling (SRS) refers to that technique in which each and every member of the population has an equal opportunity of being selected in the sample. In simple random sampling, the item gets selected in the sample depends on a matter of chance – personal bias of the researcher does not influence the selection. One of the popular and simplest methods under Simple Random Sampling is Lottery Method of Sampling in which all the items of the population are numbered on separate slips of paper of same size, shape and color and are mixed up in a box or a container before a blindfold selection is made. The desired number of slips are then selected which wholly depends on a matter of chance.

Advantage of Simple Random Sampling:
                                i.            The method is quite simple to use
                              ii.            It requires minimum knowledge of the population in advance
                            iii.            It lacks personal bias
                            iv.            It is more representative of the population as compared to judgmental sampling
                                i.            A complete list of all units/members in the whole population is needed
                              ii.            Time and cost involve can be high if the units/members are geographically scattered

b)      Systematic Sampling:  Systematic sampling is a probability sampling method where the items are chosen from a target population by selecting a random starting point and then selecting other members after a fixed ‘sampling interval’. Sampling interval is calculated by dividing the entire population size by the desired sample size. In simple words, systematic sampling can be used if a population is accurately listed or is finite. In this method, at first the list is prepared in some order and then the first item is selected randomly and subsequent items are chosen by taking every nth item from the list.

c)      Stratified Sampling: Stratified sampling is a probability sampling technique wherein the the entire population is divided into different subgroups or strata, then randomly selects the final sample proportionally from the different strata or subgroups. In stratified random sampling or stratification, the strata are formed based on members' shared characteristics or attributes such as age, gender, socioeconomic status, religion, nationality and educational attainment. Stratified random sampling is also called proportional random sampling or quota random sampling. The population within a strata or sub-groups is homogeneous with respect to the characteristics under study. Thus, in contrast to Cluster sampling, in Stratified sampling, the groups/cluster/strata are externally heterogeneous but internally homogeneous. 

d)     Cluster Sampling: Cluster sampling is a sampling method in which the entire population of the study is divided into externally homogeneous but internally heterogeneous groups called clusters. Essentially, each cluster is a mini-representation of the entire population. After identifying the clusters, certain clusters are chosen using simple random sampling as the sample of the proposed study. The population within a cluster should ideally be as heterogeneous as possible, but there should be homogeneity between clusters
Simple Random Sampling and Stratified Sampling methods are usually chosen in situations when population size is small and units are identifiable. However, cluster sampling is suitable if the population is larger.

e)      Multi-stage Sampling: Multi-stage sampling is a complex form of cluster sampling where samples are taken in different stages by dividing the population into smaller groups (or clusters) at each stage. For example, in a national level survey, at first researcher may select a few states as sample. At the second stage, within the states, a few districts may be selected and then within each district, some blocks and then some villages and so on. In short, multistage sampling divides large populations into stages to make the sampling process easy and more practical.

NON-PROBABILITY SAMPLING: It is a group of sampling techniques where the samples are collected in a way that doesn't give all the units in the population equal chances of being selected. The selection process in this method is at least partially subjective. Non-Probability sampling doesn't involve random selection at all. Non-Probability sampling can again be categorized into the following types:

a)      Purposive sampling:  It is a non-probability sample design in which the researcher purposively or deliberately selects certain units of the universe to form a sample that would represent the whole universe. In other words, Purposive sampling is the one in which the researcher chooses the sample on the basis of subject being appropriate for the study. This is used primarily when there are a limited number of people who have expertise in the area being researched. The researcher in this technique relies on his/her own judgment when choosing a sample. Purposive sampling is different from convenience sampling and is also known as judgmental, deliberate, selective, or subjective sampling.

b)      Convenience sampling: Convenience sampling (also known as grab sampling, accidental sampling, or opportunity sampling) is a type of non-probability sampling in which sample is taken from a group of people easy to contact or to reach.  This type of sampling is most useful for pilot testing.

c)      Quota sampling: Quota sampling is a non-probabilistic sampling method where the researchers divide the survey population into mutually exclusive subgroups. These subgroups are selected with respect to certain known features, traits, or interests. People in each subgroup are then selected by the researcher by fixing a quota and thus the assembled sample has the same proportions of individuals as the entire population with respect to known characteristics, traits or focused phenomenon.

d)     Snowball sampling: Also known as referrals, the sample is made up of referrals from subjects who identified other suitable subjects, usually in areas that are difficult to conduct research in. In other words, this sampling technique involves tracing rare sample through referrals from the already identified members of the sample. In short, in this method, the research participants recruit other possible participants for the study.

Difference between Probability & Non-Probability sampling
Key elements
Probability sampling
Non-Probability sampling
This group of sampling methods gives all the members of a population equal chance of being selected. Also known as Random sampling.
It is a group of sampling techniques where the samples are collected in a way that doesn't give all the units in the population equal chances of being selected. Also known as Non-random sampling.
Basis of selection
Opportunity of selection
Information about the entire population known & available
Information about the entire population is not specified and unknown
Less prone to bias and sampling errors
More prone to bias and sampling errors
Cost of procedure
Expensive, time-consuming
Less expensive, more convenient


1.   Which of the following can be considered as probability sampling?
A.     Quota sampling
B.     Lottery method of sampling
C.     Purposive sampling
D.    Convenience sampling

2.   Which of the following sampling techniques can be used to locate a sample of rare talents?
A.    Quota sampling
B.    Purposive sampling
C.    Snowball sampling
D.    Convenience sampling

3.   Which of the following cannot be considered as probability sampling?
A.     Lottery method of sampling
B.     Snowball sampling
C.     Systematic sampling
D.     Cluster sampling

4.    Which of the following techniques is used when the population is finite?
A.    Deliberate sampling
B.     Systematic sampling
C.    Area sampling
D.    Cluster sampling

5.   An alternative term for “Population” in sampling is:
A.    The sample size
B.    The Universe
C.    Cluster
D.    None of the above

6.    In which of the following sampling techniques is the ultimate size of the sample not fixed in advance?
A.    Quota sampling
B.    Cluster sampling
C.     Sequential sampling
D.     Systematic sampling

Post a Comment