Before proceeding with the sampling techniques/methods, let’s first understand what exactly sampling is. Sampling is a technique of sorting out individual representatives or a division of the population to extract statistical derivation from them as well gauge features of the entire population. For the same, there are different sampling methods. That said, the Sampling methods refer to the technique of selecting members from the population to involve in the study by using different methods.
Sampling in statistical study is of two types i.e. – probability sampling and non-probability sampling. Let’s closely review the two methods of sampling that can be implemented in any market survey.
Probability sampling is a sampling technique where a researcher sets the benchmark of selection based on a few criteria and selects members of a large population randomly. All the members bear an equal opportunity to be participating in the sample with this selection criterion. It is a conclusive type of sampling.
For example, in a community of 10,000 residents, every member will have a 1/10,000 chance of being chosen to be a part of a sample. Probability sampling is free of prejudice. Thus, removes bias in the population and provides all members a fair opportunity to get involved in the sample based on a fixed process.
Now, probability sampling also has 4 different methods under its kingdom which are as below
Simple Random Sampling: One of the choicest probability sampling methods that saves time and resources. It is a well-grounded technique of collecting information where every single member of a population is selected randomly, solely unintentionally. Each individual will have the same probability of being opted for to be a part of a sample. For example, in a society of 1000 residents, if the facilities head decides on conducting sanitization activities, it is largely possible that they would prefer picking chits out of a box. In such a case, each of the 1000 residents has an equal chance of being chosen.
Systematic Sampling: Survey creators use the systematic sampling method to select the sample members of a population at regular or systematic intervals. It needs the choice of an initial point for the sample and sample size that can be repeated at systematic intervals. This type of sampling technique has a pre-established range, and thus is the least time-consuming. For example, a researcher seeks to obtain a systematic sample of 1000 people in a population of 20,000. The researcher numbers each component of the population from 1-20000 and will select every 10th individual to be a part of the sample (Total population/ Sample Size = 20000/1000 = 20).
Cluster Sampling: Cluster sampling is a technique where the researchers segment the whole population into groups or clusters that exemplify a population. Clusters are determined in a sample on the basis of demographic like age, gender, location, education etc. This makes it quite easier to procure productive derivation from the feedback. For example, if the Indian government intends to assess the number of foreign immigrants living in the country, they can divide it into clusters based on states such as Maharashtra, Karnataka, Tamil Nadu, Andhra Pradesh, West Bengal, Uttar Pradesh, Uttarakhand etc. This way of a survey is more effective as the outcomes will be categorized orderly into states and renders insightful immigration data.
Stratified Random Sampling: It is a sampling technique in which the surveyor segments the population into smaller groups that don’t overlap but indicate the whole population. While sampling, these groups can be arranged and then pull out a sample from each group distinctively. For example, a surveyor intends to evaluate the purchasing preferences of people belonging to different annual income groups and create echelon (groups) as per the annual family income. E.g. – less than 5-6lacs–, 10-15lacs, 20-30 lacs etc. This way, the surveyor concludes the attributes of people belonging to different income groups. This will further help Marketers to assess which income groups to target and which ones to eradicate in order to form a blueprint that would yield favorable outcomes. .
(Image to be added soon)
Non-probability sampling is another sampling technique where the researcher selects members for research at random. The technique does not include a fixed or predefined selection process. This makes it quite challenging for all components of a population to have equal chances to be part of a sample. It is an exploratory type of sampling.
There are 4 types of non-probability sampling methods of which each sampling method has a specific purpose. Let’s look closely into their functionality
Convenience Sampling: This technique relies on the ease of access such as to survey buyers at a mall or passers-by on a busy street by getting in touch with the subjects. The elements of the sample are solely selected based on accessibility and not characteristically. Thus it makes for a practical method of sampling when there are time and cost restraints. For example, travel agencies generally conduct convenience sampling at a mall or public places to distribute upcoming events– they do that by handing out leaflets randomly.
Judgmental or Purposive Sampling: A judgmental sample is drawn based on the judgment of the researcher. What Researchers take into consideration is solely—the intent of the study, together with the understanding of the target audience. For example, when researchers seek to gain insight into the thought process of people willing to work in foreign countries. The parameter is: “Are you willing to work abroad…?” and those who respond with a “YES” are included in the sample.
Quota Sampling: The selection of members happens to be based on a predefined standard. In such a case, a sample is created resting upon a particular characteristic. The sample formed will have the similar attribute as that of found in the total population.
Snowball Sampling: The sampling technique is usually when the subjects are difficult to locate. Sampling method is also widely used in events where the concern is extremely sensitive and not openly discussed—says for example to collect data about HIV Aids+. For example, it will be exceptionally difficult to trace illegal immigrants. In this case, employing the snowball theory can help researchers track some divisions to interrogate and extract results.
(Image to be added soon)
Practice Problem 1
Each student in a college has a roll number. Administrator has a computer generate 202020 random identification numbers and those students are asked to participate in a survey.
What Type Of Sampling Method Is This?
Answer: Stratified random sampling
Practice Problem 2
Example A local hotel chain intends to survey its visitors one day, so they randomly sent out questionnaires that day and surveyed every visitor on the venue.
What Type of Sampling Will Be Best?
A cluster sample as it will have every member in access. Moreover it's good when each individual reflects the population as a whole.
The main advantage of probability sampling methods is that they ensure that the sample selected is indicative of the population
There are wide applications of probability sampling.
The likeliness to prejudice in sample obtained is negligible to non-existent
Probability sampling take into account vast and diverse population and the data is not fudged towards one demographic
Random sampling is considered to be a wholesome form of probability survey sampling
Systematic sampling is often used instead of random sampling
Due to minimal error, Stratified sampling is frequently incorporated probability method than random sampling
Non-probability sampling, in most situations, leads to skewed outcomes
A non-probable sample is much more useful for fundamental stages of research
Quota sampling is the quickest method of obtaining samples.
1. Why Is Sampling Used In Research?
Being time-convenient and cost-effective methods of sampling are extensively executed by researchers in market research.
Some of the popular Use of Sampling Methods by Researchers include:-
Forming the basis of any research design
Get relieved of researching the entire population
Gather functional insights
Deploying research survey software for optimum inference
Take For Example: if a pharmaceutical company would want to research the favorable effects of a drug on the country’s population (like for covid-19 potential vaccine drug), it is nearly impractical to conduct an experiment that includes everyone. In such a case, the researcher draws out a sample of people from a specific demographic and then studies them, providing the individual with the suggestive response on the drug’s behavior.
2. What Is Sampling Bias?
Sampling bias takes place when some members of a population are regularly more likely to be chosen in a sample than others. The practice is termed as bias and unjust. That said, if a sample isn't randomly or unintentionally selected, it will have a high propensity to be biased in some manner and the data may not be indicative of the population.
For Example, If the HR team selects the team lead of a department not based on the performance of activities being conducted or random chit picking but otherwise.