-SAMPLING-

Based on my reading, sampling is one of the crucial parts in a research. It features the process of selecting the participants from a larger group of people. Thus, it ensures accurate conclusion of how a group of people acts and what do they believe. 

What is a Sample?

Samples and populations

·    a group of people that the researcher has chosen using certain kind of methods to obtain information for the research.

·    population is the larger group of people whom the researcher is likely to generalize the results of the study.

·    In general, a group could be a sample in one context and population in another context.              

Defining the population

To select a sample, it is important to define a population of interest.

Target versus accessible populations

·    Target population is also known as the actual population.

·    This is the favourable type of population that researchers prefer to generalize.

·    However, researchers usually find a hard time to find one due to several limitations such as its size and the time taken. 

·    Therefore, researchers will usually generalize the results to the assessable population as it is realistic and available to be the population of the research.

Random versus non-random sampling

Two main types of research feature random sampling and non-random sampling.

Random Sampling Method

Random sampling is the method where each member of the population has the equal opportunity to be selected as the sample. There are several methods in conducting random sampling which are; simple, stratified, cluster and two-stage random samplings.

·    A simple random sample offers equal chance to each member of the population to be selected. It is preferable if the sample is in a large size. A table of random numbers is often used as the method to obtain the sample.

·    Stratified random sampling is a process of selecting the samples from the subcategories. The samples should be in the same proportion as the actual population.

·    Cluster random sampling is a process of selecting samples from the groups of subject rather than individuals.

·    Two-stage random sampling combines cluster random sampling with individual random sampling. The procedure starts with the researcher selects the samples randomly from the population. Then, the researcher selects a number of individuals from each sample. This type of sampling saves time and ensures representativeness. 

Non-random Sampling Methods

There are three methods of non-random sampling. It features systematic, convenience and purposive sampling.

·    Systematic sampling involves a detail planning of selecting sample. The researcher selects every nth from the population. This is called as random start. The other two terms that are often used in systematic sampling are sampling interval and sampling ratio. Sampling interval is the distance between each of the individual selected while sampling ratio is the proportion of the sample size and population size.

·    Convenience sampling is a process of selecting individuals who are available to be the sample for a study. This method puts the researcher at ease. However, it could lead to the issue of biasness as the sample is either strongly agrees to the topic or the other way around. This method should be avoided by the researcher, if possible. Otherwise, the researcher should repeat the procedure in order to ensure that the result is valid and reliable.

·    In purposive sampling, the researcher relies on his/her judgment in which based on the information and what he/she believes. The possibility that the judgment of the researcher may be wrong is the disadvantage of this method.

Sample Size

There are some suggestions on the size of the sample. The minimum sample size for descriptive studies is 100 individuals as compared to correlational studies, 50 individuals. On the other hand, comparative, experimental and causal studies share similar minimum sample size with 30 individuals each.

External Validity: Generalizing from a Sample

The term external validity clarifies the extent in which the results of a study can be generalised.

Population generalizability

• Population generalizability is related to the term ‘representativeness’. In one way or another, it means to what degree a sample could represent the population of interest.
• There are issues to be highlighted in ensuring representativeness of a population such as purposive or convenience samples could not represent all types of related research study.
• Furthermore, in the case of losing the select sample for more than 10 percent, it should be highlighted in the limitation of the research.
• Last but not least, it is essential to select sample from both recipient and the one who implement the treatment to ensure the representativeness of a research.

When random sampling is not feasible

When random sampling is not feasible, the replication of the research study is a wise thing to do by the researcher. The researcher should repeat the study using different group of subjects in different situation.

In the situation where the results are similar, then only the researcher could generalize the findings.

Ecological generalizability

Ecological generalizability is the degree to which the result of a study can be extended to other settings or conditions. It is important to ensure that the conditions of the settings are similar in all aspects.