A good sample should be representative of the part of the population in which we are interested, with all study participants having the same chance of being selected for the study. For example, if the population has a large ethnic subgroup, the study should use a stratified Product Sampling method. No matter the sample used and the method by which it is selected, it is important that the selected individuals are representative of the entire population.
With non-probability samples, researchers and sampling agencies are not able to randomly select participants from a random population. Non-random selection increases the likelihood that the sample has selection distortions and that the sample does not represent the population we want to study. We can never know to what extent the sample is biased, but the greater probability is more convenient if the sample is not a population compared to a random sample.
This lack of representativeness can be due to wrong selection procedures, product sampling distortions, denial of probability, non-participation in studies related to the subject matter of research or non-report.12 Most studies are carried out based on samples that do not represent the target population. For example, in the case of melanoma is available in a national or regional database and information is available on potential risk factors, it is preferable to conduct a census rather than a sampling. It is important to understand why we are sampling populations: for example, some studies are being conducted to investigate the relationship between risk factors and disease. Since we cannot study the entire population due to feasibility or cost limitations, we must select a representative sample of the interested population for observation and analysis.
It is impossible to study every person in a target group so that psychologists select samples from subgroups of the population that are likely to be representative of the target group they are interested in. The more representative the sample, the more confident the researchers are that the results can be applied to the entire target group. Researchers can identify the different types of people that make up the target populations and determine the proportions required to obtain a representative sample.
Random sampling ensures that the results of your sample approximate those of the total measured population (Shadish et al., 2002). To find the best sample size for your target group, you need results from various studies. A simple random sample allows each population unit to have the same chance of selection.
In the sample, a representative population group is selected for a study by the sampling agency. In surveys and research, the sample is a process in which a part of the population represents the entire population. Random sampling is such a method of sabotaging a sample of all units of the population with the chance to allow the generalization of the sample (Shadish, Cook & Campbell, 2002).
In the sample, units are selected from the interested population, so that each unit represents the entire population. To create a representative sample it is assumed that each intact unit represents a multitude of individuals and that a sufficient number of heterogeneous intact groups are selected as a whole to represent the whole population. When selecting a sample from a population, the sample itself is divided into parts known as sampling units or units.
This is unlikely, however, as a particular sample may represent different subsets of a particular population, and some samples may mean something different from others. It may not be possible to contact all members of a population to sample a subset of them, including those performing statistical studies. In such cases, it might be more reasonable to divide a population scattered over a large geographical region into clusters along geographical boundaries, take samples from a few clusters, and measure the units within those clusters.
One way to get a random sample from a specific person or population is to use a table of random numbers to determine which people should be included1. For example, if you have a sampling range of 1,000 people labeled 0 to 999, use a group of three-digit random number tables to select your samples. Currently, if you need to order a sample size from population size X, you can select the n-th person from a sample size N. For example, if you want a sample size of 100% of the population of 1000, choose 1000 x 100 as the 10th member of the sampling frame. If the random sample is representative of a particular population, then your three sampling tools (three different random samples) are identical or equal to the population parameter and the variability of the sampling tools is zero.
Random sampling requires a way to name a number of the total target audience and use some kind of raffle method to determine who makes up the sample. In this case, the type of sample used is by definition not random, because not all teachers in the school population have the same chance of being selected to participate. This may sound like a conclusion to draw from reading an intervention study, but researchers must be able to generalize the results of their student sample to a broader student population of interest.
The sample is a method that allows researchers to obtain information about a population from the results of a subset of that population without having to examine every individual. Researchers and sampling agencies should be aware that sampling results can be influenced by random errors, and product sampling illustrates this concept when we look at a research study aimed at estimating the prevalence of pre-malignant skin lesions as a result in individuals under the age of 18 living in a particular city that is a target population. Stratified samples are useful when researchers know the target population well enough to decide, subdivide and stratify them in a way that makes sense for research.
