Profatibility and non profatibility of sampling techniques

Data processing in markets and businesses has taken a different direction, with sampling now being commonly used. The aim of this paper is to delve deeply into the two forms of sampling: probability and non-probability dependent statistical evidence. These two types of sampling methods are further subdivided into other groups, which will be discussed further below. Probability sampling assumes that any person in a sample has an equal probability of being selected as the rest of the group. In order to create a sample, it primarily employs random sampling techniques. On the other hand, non-probability sampling is where there is the availability of cases and judgments of the interviewer. It merely uses the non-random processes. Further discussion of this is during illustration and distinguishing between the two. The preference by various companies depends on the advantages and disadvantages of both. These merits and demerits are the consideration before choosing the sampling technique to be used.

Comparison of population data across different regions has limitations in the concepts and measurements. Similarly, the difference in residence definition affects counting. Density level is one of the primary concerns in the current twenty-first century and the rate at which it has been increasing over time (Fuagier & Sargeant 2013 p.793). The primary reason for the increase in the population today is because of the rise in fertility and reduction in the mortality rate. This growth means more children are born, and fewer deaths reduction leading to the addition of humans. However, the price of this increase in the population varies with regions or from one society to another. Each of the countries in the world thus has different patterns and transitions. A recent study comparing China and the USA found that both are undergoing spatial and demographic transformations which not only in the two countries, but also in other nations (Weir & Cockerham 2015, p. 1368).

The sample size is a significant undertaking in a research process. In the case of calculations, for instance, they should be appropriately counter-checked before the final decisions are made (Marshall, Cardin & Poddar 2013, p.15). A significant reason behind the use of sample sizes is because it gives a precise detail which most often brings complications and further illustrations of the method of an example. Another instance of a merit involved in sample size is where the research suggests a few subjects. There will be a waste of time because a realistic solution is not necessarily reached by a single random sample (Marshall, Cardin & Poddar 2013, p.18). Too large study leads to waste of resources, time included. However, providing a sample size is not only giving one number from a set of tables. It involves several processes in different stages. This shows that the sample size in quantitative research in business is vital in as much as it indicates the resources that are useful and needed it also helps the researchers can determine the design issues critically (Machin, Campbell & Tan 2011, n.p).

. PROBABILITY SAMPLING

As mentioned earlier in the introduction, probability sampling based on the available observations or sampling distribution. Marahal (1996) in his study describes probability sampling as the techniques used in quantitative studies. He uses quantitative sampling as the central theme of his research (p. 1). He further reinstates that the primary objective is to draw a quantitative sampling from a population so that a general conclusion is reached from the sample of the people. Researchers select an appropriate example, depending on the aim of the study to be undertaken. The sample size determines the optimum number required to allow correct inferences to be made on a population. The smaller the size of the sample, the larger the chance of a random sampling error, but since when the sampling error goes up, the square root of the example, reduces and there is only a small knowledge gained from the various samples (Marshall 2016, p. 525).

EXAMPLE OF PROBABILITY SAMPLING

Figure 1: Probability sampling (Marshall 2016, p. 526)

From the chart above, an illustration of probability sampling is shown three students are needed for a group of twelve. Each element is given a random number from the given data. The highest assigned name for each row is chosen (56, 92 and 63 in the first, second and third row). Further analysis of the selected samples is done.

TYPES OS PROBABILITY SAMPLING

Probability sampling is used in qualitative studies (QUAL) and mostly involves selection of units or individuals based on the aim of the research. It mainly requires a range of a significant number of the units or individual from a population, and the probability of choosing an individual from the group can be determined. Probability sampling can be classified into three categories (Moshagen, Hilbig & Erdfelder 2014, n.p.).

Random sampling-this occurs where every individual in a population has an equal chance as the other to be included.

Stratified sampling-this is where there is a division into subgroups called strata so that the researcher picks the units from the levels.

Cluster sampling- this is where there exists a group within a population.

Random Sampling

This is a type of probability sampling, where each unit or an individual in an existing population, while conducting research has the same chance of being included in the sample. However, the likelihood of a group in a case to be picked is not necessarily affected by other units from the population. Simple random selection can be made in several ways. This includes drawing numbers or values out in a cube or preferably using a computer programme to randomly select numbers (Weir & Cockerham 2015, p. 1359). This is thus the most well-known type of sampling technique and widely used.

EXAMPLE 2

Figure 2: Simple random sampling (Moshagen, Hilbig & Erdfelder 2014, n.p.)

From the picture shown above, the clients are randomly chosen. This is the most natural form of sampling. A sample is drawn, and a decision is reached on the number of clients to be determined, let’s say 100, but the total number of clients is 1000. The sampling fraction is n/N. This becomes 100/1000 which is 10%. Now what remains for the surveyor is to draw a sample because he or she has several options to choose from.

Stratified Sampling

This is a method of sampling which uses both stratified and random sampling. This can be explained regarding the interest of a researcher where if he or she is interested in randomly choosing a case then a sample has to be picked from the representative of the population depending on the aim of the research. A limitation only strikes when the researcher wants representation in various subgroups (Moshagen, Hilbig & Erdfelder 2014, n.p.). An example, in this case, would be a situation in which a researcher wants a sample in the stratified method of sampling in a college of ladies and gentlemen in the class. The sampling will be a situation of dividing into two groups of males and females. The researcher further goes into selecting a random sample from the two groups formed.

EXAMPLE 3

The figure 3 below shows an example of stratified sampling that was done with a list of clients divided into three groups; Caucasians, African-American, Hispanic-American. A pure random is done and assuming that n=100, then the probability of getting 10-15 person from each team is high. The sample is now stratified, and the determination of the number of people the researcher wants in each group is done. This is a case of stratified sampling.

Figure 3: Random samples of n/N (Wybo, Robert & Leger 2015, p.5)

Cluster Sampling

This happens when the researcher is in need of producing a less costly and efficient regarding time probability sample. The researcher prefers to go clusters or groups in a population instead is individual unit samples. An example of this sampling technique is sampling of the schools.

ADVANTAGES OF PROBABILITY SAMPLING

Every sampling technique comes with various advantages and disadvantages. This can be seen in that in some businesses choosing of the best sampling technique is done before conducting research (Wybo, Robert & Leger 2015, p.8). This will give a room the researchers to ensure that a proper investigation is conducted. The merits of this kind of technique are based on the categories, meaning that every type comes with its advantage:

Cost Effective

This is through random sampling whereby after the samples have been chosen in a particular group or population, the percentage of the process done is averagely fifty percent. This saves time as well as cost (Wybo, Robert & Leger 2015, p.8).

Involves Lesser Degree of Judgement

The process of random sampling is more efficient and accurate given that it just involves giving out a number of the population. This thus does not include a lot of judgments (Wybo, Robert & Leger 2015, p.8).

Comparatively Easier Way of Sampling

It does not involve the long or complicated process of sampling.

Cluster sampling is the preferred kind because of its convenience and ease of use. It involves selecting the most accessible subjects, and it is the most affordable kind of probability sampling technique. Most researchers afford this type of technique.

A simple random sampling is advantageous as it involves the creation of samples which represents a more significant percentage of the population being of the research (Fuagier & Sargeant 2013 p.796). The merit comes in because it gives clear and vivid information on the research results.

DISADVANTAGES OF NON PROBABILITY SAMPLING

Chances of Selection of Particular Cases of Samples

If for instance, a researcher collecting samples on family children, the probability of choosing only the eldest sons from different families is high. This will mean that only research will be based on the earliest sons of different families.

Monotonous work

A surveyor does repetition of the job when the research involves taking or giving out numbers. This becomes monotonous to the surveyor.

NON-PROBABILITY SAMPLING

This is also referred to as purposive sampling. It occurs where individual units are selected based on the aim of the research. It does not give every individual in a population a chance to be selected. In most cases, it creates its control from the conclusion of the one doing the research. The situations chosen for study are based on the availability and the researcher’s conclusion. The main aim of the purposive sampling is to make comparisons between different types of situations in research (Mahmud 2017, n.p.). In as much as non-probability sampling has common objective, it is classified into three categories;

Sequential sampling

Sampling cases which are unique.

Sampling for comparability

Use of multiple purposive techniques for sampling. Intensity sampling.

Sampling Cases which are Special

Include cases which are unique and have been the primary focus of quantitative research. This can be attributed to anthropology and sociology. An example of this kind of sampling technique is the revelatory case sampling, where entry to a particular case is sought whereby the phenomenon had not been accessed in the past (Fuagier & Sargeant 2013 p.796). However, these cases are very hard to come across and often very hard, but yield precious information of unstudied instances. Further, an example of a revelatory case sampling is a study of a plantation farm which consisted of the low-income families. The finding showed how the unique environment provided a different kind of the surroundings to the families. This is a case of exceptional case sampling.

Sequential Sampling

This is a broad category which involves gradual or slow selection of samples for research. The technique is divided into four categories:

Theoretical sampling- is a technique in which a researcher explores on one instance of a phenomenon at a time for elaborations after scrutiny and come up with manifestations. The author gives an example of this technique as that of “awareness of dying.” They took different sites of research, and each section provided unique awareness of death.

Snowball sampling-this is where the existing study subjects recruit future subjects and are mostly faced with a limitation of bias.

Opportunistic sampling- this is where decisions are met by the researchers while collecting data.

Confirming and disconfirming cases-this kind of sequential typing occurs when the completion of the process of collecting data and analysis had already been done.

Sampling Using Multiple Purpose Techniques

This is where several purposive techniques are combined and is applicable when two or more sampling strategies are used when picking cases. Complexities in sampling have been a significant setback, but the use of this technique solves the technicalities involved in sampling. An example of this method is about the abuse and oppression of women (Marshall 2016, p.523). The researcher used all the four types of sequential sampling and selection was done for every group.

Strategies in Non-probability technique

In non-probability sampling, there exist several types of procedures (Cooper & Schindler 2006, n.p). One of them being sampling for comparability or representativeness. This is where a surveyor selects a nonprobability sample that belongs to a group of cases. The closeness should be as high as possible. Another goal of this type of sampling is the comparison of samples of different kinds of situations. Another strategy in purposive sampling is the sampling the unique or unusual cases. This is introduced when the primary focus of the research is an individual or particular group of cases.

Quota sampling has also been found to be one of the examples or categories of nonprobability sampling (Marshall 2016, p.523). It involves the researcher or the surveyor ensuring that each subject gets an equal or a proportionate sample of the quota. In any research study, reliability and validity should be taken into consideration. The extent of production of results of similar results under constant conditions is called reliability.

Sequential sampling being part of the non-probability technique employs a slow selection based on how the research questions are relevant. Purposive sampling also encompasses sampling by use of multiple purposive methods in the same study. Nevertheless, purposive sampling has up to six procedures based on achieving comparability. These are maximum variation, deviant case sampling, homogeneous sampling, reputational sampling and intensity sampling. (Mahmud 2017, n.p). Comparison and contrasts are the significant concerns in the nonprobability technique of sampling. For instance, extreme and deviant cases provide comparisons by showing the different ideas with other examples in research.

(Mahmud 2017, n.p) found that quantitative sampling’s main aim is to absorb detailed data and the pattern of the data information to be collected. He further puts forward that it may take ages to research quantitative sampling since it requires different concepts and ideas of the subject. A comparison of the quantitative research showed that this kind of sampling g is more expensive to quantitative research. In this study, respondents have a freedom of choosing from the research questions. Data is collected and put in the form of a draft, and this will mean that changing of the information can be done in the process.

ADVANTAGES OF NON PROBABILITY SAMPLING TECHNIQUE

Nonprobability sampling technique is active when the conduction of probability sampling is not practical. This calls for this kind of sampling technique.

This kind of sampling technique is cost effective and time efficient.

DISADVANTAGES OF NON PROBABILITY SAMPLING TECHNIQUE

It poses difficulties when determining the sampling variability.

When a comparison is made with the probability sampling, this kind of sampling gives a little generalization of the research being conducted.

Representation is that of a higher portion.

CONCLUSION AND RECOMMENDATION

Sampling for quantitative research in business is an area of confusion for researchers and it broadly relates to the misunderstanding in the quantitative approach where proper understandings of the human issues which are tricky are more critical than generalizing results. From the above report, it is probably clear that every sampling requires not only critical analysis of the technique to be used in sampling, but also a consideration of the limitations it is involved in. Researchers and surveyors in business should emulate the need to choose the right quantitative business technique. Probability sampling does not involve many complexities, and it is suitable for sampling. Nonprobability sampling, on the other hand, is good for sampling that involves many complications, and this calls for researchers and surveyors to choose wisely depending on the type of business research and survey they are carrying out.

References

Cooper, D.R., Schindler, P.S. and Sun, J., 2006. Business research methods (Vol. 9). New York: McGraw-Hill Irwin.

Faugier, J. and Sargeant, M., 2013. Sampling hard to reach populations. Journal of advanced nursing, 26(4), pp.790-797.

Machin, D., Campbell, M.J., Tan, S.B. and Tan, S.H., 2011. Sample size tables for clinical studies. John Wiley & Sons.

Marshall, B., Cardon, P., Poddar, A. and Fontenot, R., 2013. Does sample size matter in qualitative research?: A review of qualitative interviews in IS research. Journal of Computer Information Systems, 54(1), pp.11-22.

Moshagen, M., Hilbig, B.E., Erdfelder, E., and Moritz, A., 2014. An experimental validation method for questioning techniques that assess sensitive issues. Experimental Psychology

Weir, B.S. and Cockerham, C.C., 2015. Estimating F‐statistics for the analysis of population structure. evolution, 38(6), pp.1358-1370

Wybo, M., Robert, J. and Léger, P.M., 2015. An optimization model of the business applications selection process. CashMachin, D., Campbell, M.J., Tan, S.B. and Tan, S.H., 2011. Sample size tables for clinical studies. John Wiley & Sons.ier du GReSI no, 5, pp.08-08.

Appendix

Figure 4: Population Distribution in New York State (Wybo, Robert & Leger 2015, p.10)

The figure 4 above shows population distribution in New York State. Sampling will involve moving from town to town to collect data. Samples are selected depending on the type of sampling technique employed. Moving from town to town will help in saving cost rather than walking in the whole of New York State to collect the population data. A proper business technique should be chosen.