Sampling Methods

 Sampling Methods

 

 

Nonprobability sampling 

Nonprobability sampling refers to the sampling technique in which an individual probability of getting selected is unknown due to uncertainty of selection. In Nonprobability sampling it is unknow whether the sample is representing the population or not. But the main aim is not the generalization of the population. In this type of sampling technique researchers selects a sample based on the subjective judgment of the researcher instead of random selection. It is often used for qualitative sampling.

Types of non-probability sampling

Here are some types of non-probability sampling methods:

·       Convenience sampling

·       Consecutive sampling

·       Quota sampling

·       Judgmental or Purposive sampling

·       Snowball sampling

Judgmental Sampling

 

Also known as purposive sampling which in non-probability sampling technique where sample members are chosen on the basis of researcher’s judgement and knowledge.

This sampling gives highly accurate with minimum margin of error output as the researcher’s knowledge is instrumental. The knowledge of the researcher is primary in this sampling process as the members of the sample are not randomly chosen.

 

When to execute Judgmental Sampling?

Purposive sampling is more effective in situations where there are only a restricted number of people possessing the qualities that the researcher expects. This technique is usually preferred because other techniques will be more time-consuming and when the researcher has confidence in their knowledge to select a sample.

Judgemental sampling is used where the authorities involved would prefer relying on their knowledge and not on other sampling methods and also where there are time constraints.

Purposive sampling is usually used in situations where the target population comprises highly intellectual individuals who cannot be chosen by using any other probability or non-probability sampling technique.

 

Judgmental Sampling Advantages

1. Almost real-time results as a poll or survey can be carried out using this technique as the members possess the understanding and required knowledge of the subject.

2. Target audience can be reached directly as it completely depends on the researcher’s preferences and thus desired results can be produced.

3. Minimum time required for execution as there are no barriers due to which the sample selection becomes more convenient.

 

Consecutive Sampling

Consecutive Sampling is that the process of conducting research including all the people that meet the inclusion criteria and are conveniently available, as a part of the sample. For this system the researchers conduct research one after the opposite until they reach a conclusive result. Over here, the sample is selected based on their easy availability after that research is conducted, results are obtained and analysed and then the researcher moves on to the next sample.

 

A pre -requisite for conducting this research is that outlining the requirements of people to be put within the population then selecting the sample supported convenience. Once this requirement is passed, each of the potential respondent is evaluated to see if they meet the listed requirement. Once they meet the checklist, they're included within the sample population to hold out the the research. It is very crucial that the researcher makes sure that the need list covers all the aspects.

 

Example

 

·       A shoe brand decides to launch a replacement shoe design that they decide to conduct a search to check the market acceptance for that specific design.

·       They then culminate the results obtained from each of their outlets and analyse it to succeed in a conclusion for that specific design.

·       Once analyzing, they find the results inconclusive as many aren't sure about the color coding of the new design while others find it exciting. This prevents them from reach a firm decision of whether to launch the product or not and this will matter a lot.

·       So, they modify the survey inquiries to get obviate any longer ambiguity to arrive upon concise and clear results and include a neighborhood that permits respondents to suggest changes that they feel should be made to form the merchandise more suitable for the market.

·       The brand conducts the survey again and this point, find that about 65% of the respondents feel that the merchandise are going to be well accepted by the buyers if launched at the earliest. Others suggest changes that the brand can make to convince customers of its necessity and make it more appealing and more likable.

 

It’s clear that within the second case, the brand was ready to arrive upon decisive results as a results of learning from their first analysis which was conducted earlier

Benefits

·       Cost effective: This sampling method doesn't require the organizations to rent separate professionals to gather research data. The researcher also saves money by avoiding hiring agencies that find them suitable respondents for the study. This allows the researchers to make cut in cost.

·       Less time consuming: Since the sample is chosen on the idea of convenience, not much time is wasted . This allows the researchers to collect data fast and leaves sufficient time for analysis.

·       Valuable: Consecutive sampling is that the most precious and usable method of sampling because it guarantees results . The researcher can move from one sample to another to satisfy the research purpose and verifying previously obtained data. This ensures both, value for money and effort for building best product.

 

Downsides

 Slice bias Although the tactic of sample selection could also be a lower quantum time consuming and reduces sweats, it can impact quality of the results attained. For illustration within the below case, one can argue that the callers of the shoe brand outlets do not represent the opinions of all target population. Also, there are chances that a specific section of the the population did not find representation within the sample. For illustration It's possible that the study didn't cover persons progressed above 60 just because they didn’t be to visit the brand outlet during the timeframe within which the study was conducted.

 

 Intentional Slice

 Intentional slice represents a covey of colourful non-probability slice ways. Also appertained to as hypercritical, picky or private slice, intentional slice relies on the judgement of the experimenter when it involves opting the units (e.g., people, cases/ organisations, events, pieces of data) that are to be studied. Generally, the sample being delved is kind of small, especially as compared with probability slice ways.

Unlike the multitudinous slice ways which can be used under probability slice (e.g., simple slice, representative slice, etc.), the thing of intentional slice is not to aimlessly elect units from a population to make a sample with the intention of creating generalisations (i.e., statistical consequences) from that sample to the population of interest ( see the composition Probability slice). This is the overall intent of exploration that is guided by a quantitative exploration design.

 

 The main thing of intentional slice is to concentrate on characteristics of a population that are of interest, which can best enable you to answer your exploration questions. The sample being studied is not representative of the population, piecemeal from experimenters pursuing qualitative or mixed styles exploration designs, this is frequently frequently not considered to be a weakness. Rather, it's a choice, the end of which varies counting on the kind of proposing slice fashion that is used. For illustration, in homogeneous slice, units are named supported their having analogous characteristics because similar characteristics are of interested to the experimenter. By discrepancy, critical case slice is generally employed in exploratory, qualitative exploration so on assess whether the miracle of interest indeed exists (amongst other reasons).

 

 During a qualitative or mixed styles exploration design, relatively one kind of intentional slice fashion could indeed be used. For illustration, critical case slice could indeed be used to probe whether a miracle is worth probing further, before espousing a maximum variation slice fashion is employed to develop a wider picture of the miracle. We explain the varied pretensions of these kinds of intentional slice fashion within posterior section.

 Advantages and disadvantages of intentional slice

 

 Advantages of intentional slice-

 There are an honest range of qualitative exploration designs that experimenters can draw on. Achieving the pretensions of similar qualitative exploration designs requires differing types of slice strategy and slice fashion. One of the foremost benefits of the intentional slice is that the large choice of slice ways which may be used across similar qualitative exploration designs; intentional slice ways that range from homogeneous slice through to critical case slice, expert slice, and more.

 Whilst the colorful intentional slice ways each have different pretensions, they go to supply experimenters with the defence to form generalizations from the sample that is being studied, whether similar generalisations are theoretical, logical, and/ or logical in nature. Still, since each of these kinds of intentional slice differs in terms of the character and skill to make generalizations, you need to read the papers on each of those intentional slice ways to know their relative advantages.

 Qualitative exploration designs can involve multiple phases, with each phase structure on the former bone. In similar cases, differing types of slice fashion could also be needed at each phase. Intentional slice is salutary in these cases because it provides a good range of non-probability slice ways for the experimenter to draw on. For illustration, critical case slice could also be used to probe whether a miracle is worth probing further, before espousing an expert slice approach to feel at specific issues further.

 

 Disadvantages of intentional slice-

 Intentional samples, anyhow of the kind of intentional slice used, are frequently largely susceptible to experimenter bias. The idea that an intentional sample has been created supported the judgement of the experimenter is not an honest defence when it involves easing possible experimenter impulses, especially in comparison with probability slice ways that are designed to gauge back similar impulses. Still, this judgemental, private element of purpose slice is simply a serious disadvantage when similar judgements are ill-conceived or inadequately considered; that is, where judgements have not been grounded on clear criteria, whether a theoretical frame, expert elicitation, or another accepted criterion.

 

Quota Sampling

Definition

Quota sampling may be a sampling methodology whereby information is collected from a consistent cluster. It involves a ballroom dancing method wherever 2 variables may be accustomed filter info from the population. It will simply be administered and helps in fast comparison.

In Depth

Quota sampling may be an easy however effective thanks to do analysis within the initial phases. From the population, the investigator might choose 2 variables to review a few explicit clusters. He might use gender likewise as financial gain level or the education level for the aim of analysis. The investigator might additionally add alternative sub-points to the info set in step with the necessities of the analysis. In a quota sampling there's a non-random sample choice taken; however it's done from one class that some researchers feel may be unreliable. The researchers run the chance of bias. Interviewers can be tempted to interview those people on the road United Nations agency seem most useful in filling the shape or sample people United Nations agency could contradict them or others acknowledged to them simply to satisfy the target set of audience. Quota sampling is employed once the corporate is brief of your time or the budget of the one that is researching on the subject is proscribed.  it's a simple method to hold out and decipher info once the sampling is completed. It additionally improves the illustration of any explicit cluster inside the population thereby guaranteeing that these teams aren't over-represented.

 

Convenience Sampling

Definition

Convenience sampling is outlined as a technique adopted by researchers wherever they collect research knowledge from a handily out their pool of respondents. it's the foremost normally used sampling technique as it’s improbably prompt, uncomplicated, and economical. In several cases, members are promptly approachable to be a locality of the sample.

In Depth: -

Convenience sampling is applied by brands and organizations to live their perception of their image within the market. knowledge is collected from potential customers to grasp specific problems or manage opinions of a new launched product. In some cases, it's the sole on the market choice. as an example, a college student functioning on a project and needs to grasp the common consumption of soda on field on a Fri night can most presumably decision his/her classmates and friends and raise what number cans of soda they consume. Or might attend a celebration near and conduct a straightforward survey. there's perpetually an opportunity that the willy-nilly elect population might not accurately represent the population of interest, therefore increasing the possibilities of bias.

A basic example of a convenience sampling technique is once firms distribute their promotional pamphlets and lift queries at a mall or on a packed street with randomly elect participants. Businesses use this sampling technique to assemble data to agitate essential issues arising from the market. They together use it once assortment feedback several specific feature or a freshly launched product from the sample created. During the initial stages of survey analysis, researchers generally create convenience sampling as it’s quick and easy to deliver results. although many statisticians avoid implementing this technique, it is very necessary in things where you propose to urge insights in associate degree extremely shorter quantity or whereas not finance associate degree excessive quantity of money. For instance, a commercialism student should get feedback on the “scope of content commercialism in 2020.”

 

Probability sampling 

Types of probability sampling

Here are some types of probability sampling methods:

•       Simple Random Sampling

•       Clustered Sampling

•       Systematic Sampling

•       Stratified Sampling

 

Simple Random Sampling

 

Simple random sampling is a probability sampling technique. It means that every member of the subset has an equal probability of being chosen. This technique is regarded as an unbiased representation of the group. Random portions of the group are considered while doing this technique of sampling.

Let’s consider an example where there are 100 students who are studying in a same class, have taken a common exam. Then a random selection of 10 students will represent the entire student population.

This method is most of the times used when we do not have much information of the population which is being considered.

It is usually used for larger populations and hence, it is important to ensure a sample size that is large enough to fittingly represent this population.

The advantages of using the technique of Simple Random Sampling include that it is considered a fair way to select a sample from a bigger population since every member has an equal probability of being selected. Another advantage it includes is that this technique is easy to carry out and it is easy.

If this technique is implemented properly, it is regarded as the best sampling method for ensuring both internal and external validity. But sometimes it is not practical and is expensive to implement, depending on the size of the population to be studied.

 

The process of simple random sampling include:

1.      The first and foremost thing to do is to define the population size we are working with.

2.      Then we have to assign serial numbers to the entire population. Those numbers will act as ID numbers.

3.      Now comes the step of deciding the sample size number which is needed.

4.      Then we need to select our sample by randomly selecting any number between the ranges of the defined population size.

a.      There are different methods through which it can be achieved. They are a) The Lottery method.

b)      Random number tables.

c)      Software packages.

 

 

Systematic sampling

Systematic sampling is a probability-based method which has some advantages as well as disadvantages. Systematic sampling is a probability sampling method in which a random sample, with a fixed periodic interval, is selected from a larger population. This fixed periodic interval, called the sampling interval, is calculated by dividing the population size by the desired sample size. In a systematic sample, chosen data is evenly distributed. 

If our population of interest is in ascending or descending order, we should make use of systematic sampling since it will include member from both the bottom as well as top ends of the population.

For example, if we are sampling from a list of individuals who are ordered by age, then the systematic sampling will result in a population which is obtained from the entire age spectrum. If instead, we use simple random sampling, it might be possible that we would end up with only younger people or older people.

One thing which should be noted is that we should not make use of systematic sampling if our population is ordered in a cyclical or periodical manner. If it is so, then the resulting sample cannot be guaranteed to be representative.

Advantages:

·       Some of the advantages of this sampling include the elimination of the phenomenon of clustered selection and also a low probability of contaminated data.

· The second advantage includes that it is Easy to Execute and simple to Understand.

·       Another advantage is that it reduces the risk of favouritism as researchers have no control over who gets selected for systematic sampling, which means that it provides a low risk of data manipulation during the work collection process and at the same time, it keeps the sampling work highly productive with a negligible risk of error.

 Disadvantages:

•  Some of the disadvantages of this sampling technique include over representation or under-representation of particular patterns and also a greater risk of data manipulation is there.

·       Also, there is a greater Risk of Data Manipulation with systematic sampling because researchers might be able to construct their systems to increase the likelihood of achieving a targeted outcome rather than letting the random data produce a representative answer. Any resulting statistics could not be trusted.

·       Systematic sampling is less random than a simple random sampling effort: 

If randomness is our top priority for research, then systematic sampling is not the best option to choose. Though it takes less time and is not as tedious as other methods of data collection, there still is a predictable nature to its efforts which can influence the end results.

 

Stratified Sampling

Definition

Stratified sampling is a type of system in which the total population is divided into lower groups or strata to complete the sampling process. The strata are formed grounded on some common characteristics in the population data. After dividing the population into strata, the experimenter aimlessly selects the sample proportionally.

 In Depth

Stratified sampling is a common sampling technique used by experimenters when trying to draw conclusions from different sub-groups or strata. The strata or sub-groups should be different, and the data shouldn't lap. While using stratified sampling, the experimenter should use simple probability slice. The population is divided into colourful groups similar as age, gender, nation, job profile, educational position etc. Stratified slice is used when the experimenter wants to understand the being relationship between two groups.

 The researcher can represent indeed the lowest sub-group in the population. There are two types of stratified slice – one is commensurate stratified arbitrary sampling, and another is disproportionate stratified arbitrary slice. In the commensurate arbitrary sampling, each stratum would have the same sampling bit. For illustration, you have three sub-groups with a population size of 150, 200, 250 subjects in each group independently. Now, to make it commensurate, the experimenter uses one specific bit or a chance to be applied on its groups of population. The sample for first group is 150 *0.5 = 75, 200 *0.5 = 100 and 250 *0.5 = 125. Then the constant factor is the proportion portion for each population subset.

The only difference is the slice bit in the disproportionate stratified sampling fashion. The experimenter could use different fragments for colourful groups depending on the type of exploration or conclusion he wants to decide from the population. The only disadvantage to that's the fact that if the experimenter lays too important emphasis on one group, the result could be disposed.

 

In stratified sampling, strata are used for dividing the population, which are often based on demographic characteristics such as gender, socioeconomic status, or race. In one stratum every unit of population is placed.

sampling units- clusters where we go from higher-level to lower-level at each stage.

primary sampling units (PSU)- refers to dividing and selecting the population clusters. It is the first stage.

secondary sampling units (SSU)- PSU are further divided into clusters.

 ultimate sampling units (USUs)- after continuing for multiple stages we reach this stage where we get our final sample.

 

Advantages

·       Stratified sampling is superior to simple sampling as a result of the method of stratifying reduces sampling error and ensures a larger level of illustration. Thanks to the selection of stratified sampling adequate illustration of all subgroups is ensured.

·       When there's homogeneity inside strata and nonuniformity between strata, the estimates are as precise (or even additional precise) like the employment of easy sampling.

·       Cost effective when it comes to geographically dispersed population.

·       Flexible as it can vary sampling methods between stages based on what’s appropriate.

·       No need of sampling frame for the target population.

 

 Disadvantages

·       The need to be ready to simply distinguish between strata within the sample frame might produce difficulties in sensible levels.

·       The choice of representative sampling technique adds sure complexness to the analysis arrange

 

Cluster Sampling

Definition

in this type of sampling the population is divide into multiple groups which allows the researchers to collect data by bifurcating the data into small, more productive groups. Then random groups are selected with a simple random or systematic random sampling technique for data collection and data analysis. Then, researchers analyse a sample that consist of multiple parameters such as background of the population, habits, demographics, and any other attributes. This type of sampling method is used when the population groups are similar yet internally diverse form a statistical population.

Example: examine a situation where a mobile company is looking to survey the performance of smartphones across India. They can divide the whole country's population into cities (clusters) and select more towns with the highest population which also screen out those using mobile devices.

Types of cluster sampling

There are two methods to classify the sampling technique. The first one depends on the number of stages followed to obtain the cluster sample.

Other way is the representation of the groups. In numerous cases, sampling by clusters takes place over multiple stages.

Single-stage cluster sampling: 

In single stage cluster sampling technique, sampling is done just once. An example of single-stage cluster sampling –

Two-stage cluster sampling: 

In two-stage cluster sampling instead of selecting all the elements of a cluster, only some of members are selected from each team by simple random sampling.

Multistage Sampling: -

It is the cluster sampling where a sample from population is drawn at every stage by using smaller and smaller groups. It is generally used in national surveys as it is effective in collecting data from large, geographically spread group.

 

Single-stage vs multistage sampling

In single stage the populations are divided into units and then a sample is selected directly by collecting data from everyone in the selected units.

In multistage sampling, sampling is done in several steps by forming number of smaller groups. This also helps in hierarchical groupings by creating a less expensive sample.

We can use either non- probability or probability sampling methods in multi-stage and single-stage sampling.

Probability sampling methods should be used for stronger statistical inferences for generalizability and external validity.

 

Cluster sampling Advantages & Disadvantage

 

There are advantages of using cluster sampling.

· Consumes less time and cost: cluster sampling of geographically divided groups requires less work, time, and cost. It’s a highly economical method.

· Convenient access: Researchers can choose larger samples with this sampling technique, and that’ll increase accessibility to various clusters.

· Data accuracy: In cluster sampling an analyst is sampling a large sample in every cluster, so loss of accuracy in data for everyone is counterbalanced.

· Ease of implementation: Cluster sampling enables data from various groups. Which allow researchers and analyst to implement it in practical situations more easily as compared to other probability sampling methods.

 

Cluster vs stratified sampling

In stratified sampling and cluster sampling, we divide the population into groups that are mutually exclusive and exhaustive.

In cluster sampling, the population is divided into clusters, which are usually based on geography (for ex. States/cities) or organization (for ex. Schools/universities). 

In single-stage cluster sampling, we randomly select some of the clusters for our sample and collect data from everyone within those clusters in one stage.

 

 Written By:

• Sanket Jagdale 
• Niraj Jangale
• Ayush Jiwatode 
• Jaai Joshi 
• Piyush Kakade