Sampling

Sampling

Another important aspect in designing a research process involves deciding who are the subjects to be studied. The object is to be able to generalize the subjects studied to others with similar conditions or characteristics.

Idea is to explain how a small representation of subjects can be used with confidence to make predictions about the larger population.

What is sampling?

In research, sampling refers to the process of selection of a smaller group of participants from the population of interest. While it would be ideal for the entire population you are researching to take part in your study, logistically this may not be feasible. Therefore, by researching a smaller and representative group obtained from your population of interest, we are able generalize the findings back to and make inferences about the whole population.

The larger group of interest to which the research results are being applied to is called the population. A population is defined as set of persons, objects or events that meet a specified set of criteria. For examples, if we are interested in studying effects on a certain drug in the treatment of diabetes, the population of interest would be all people in the world who have diabetes. However, it is not possible to test every person in the world with diabetes. Working with small groups of people is more economical, time efficient and allows for better control of the subjects. Therefore, through the process of sampling, researchers choose a subgroup of the population, called sample. This sample serves as the reference group allowing to make estimations about the entire population.

Populations are not limited to human subjects only, it can include people, places, organizations, objectives, animals, days or any other unit of interest, depending on the research involved. [1]

Inclusion and Exclusion Criteria

In selecting the target population, the researcher must first define the selection criteria that will decide which subjects can be chosen and which subjects cannot be chosen.[1] [2]

· Inclusion criteria describes the primary characteristics of the target population that will allow them to qualify as a subject. Common inclusion criteria can be demographic, clinical, or geographic in nature.

For example: You are running a clinical trial for testing new treatment drug for patients with type I diabetes, following inclusion criteria apply

· 20 to 75 years of age

· Have been on insulin pump for over a year.

· Willing to return for required follow-up visits for up to a year from the start of the treatment.

· Exclusion criteria describes the characteristics of the target population that will preclude subjects from participating in the study. Could be age, gender, ethical considerations, etc.

For example: In the clinical trial for testing new treatment drug for patients with type I diabetes, following exclusion criteria apply:

· Cannot have Type II diabetes

· Patient has history of pancreatic surgery

· Patient is cognitively challenged

· Patient had major surgery 3 months prior to enrollment.

A) Simple Random sampling: This is one of the easiest and most convenient form of sampling. Samples are randomly collected from the target population to represent the entire population, hence giving each subject an equal probability of being chosen. Simple methods like lotteries, random draws, surveys or pulling samples from directories of the organizations are the usual ways to collect the sample to represent the data. For example, in the factory of 500 workers, 50 workers were chosen randomly by lottery to participate in customer service survey.

B) Systematic Sampling: In this type of sampling technique, the subjects are selected from the target population by arranging them in some kind of order or interval e.g. arranging them in alphabetical order or choosing every 10^th person from the list of the applicable candidates. This is still a type of random sample but much easier and less time-consuming way to conduct then just simple random sampling.

C)Stratified Random Sampling: This type of probability sampling technique is also known as proportional random sampling or quota random sampling. It involves identifying relevant characteristics of the target population and dividing members of the population into homogenous and non-overlapping groups or strata based on these identified characteristics. For e.g., dividing population based on level of education, age, gender, income or demographics.

For instance, a researcher might want to know the correlation between retirement age and income they could use stratified random sampling to divide the population into strata and take a random sample from it.

D)Cluster Sampling: This type of probability sampling is particularly helpful in research studies involving large and geographically dispersed population. For examples trying to study outpatient physical therapy trends across the state. It is impossible to collect the sample from the entire state, so researchers will use pre-existing units such as counties and hospitals and choose samples from the hospitals located in those counties.

Non-probability sampling

They type of sampling is based on convenience. The researcher decided the criteria of the selection based on the needs of the study and the subjects are chosen in a non-random way. It is a much easier process of sampling, but the drawback is that the results cannot be applied to entire population. [1][4]

The non-probability sampling is often used in qualitative and exploratory research.[1] [4]

The non-probability sampling techniques can also be categorized into 4 categories:

A) Convenience Sampling: it is the most common form of non-probability sampling, also called accidental sampling. Subjects are chosen based on availability or easy accessibility of the subjects.

B) Voluntary Response Sampling: A voluntary response sampling is a type of sampling made of subjects who have voluntarily chosen to participate in the study or project. The researcher puts out a request for members of a population to join the sample via surveys or advertisements and people decide whether to be in the sample or not.

C) Purposive Sampling: In this type of sampling, the researcher hand picks the subjects based on specific criteria or their personal judgement. This can be done through chart reviews or interview. Here, the entire sampling process depends on the researcher’s judgment and knowledge of the context, if chosen wisely the researcher can filter out lot of subjects that do not fit the criteria.

D) Snowball Sampling: this technique is used then subjects with specific characteristics are hard to locate. This samples are chosen by referral from other participants. In first stage few subjects are chosen based on the required criteria. In the second stage, these subjects are asked to identify or refer other subjects who fit the same criteria. This process of “snowballing” is continued till the adequate number is obtained.

[1] From Portney L and Watkins M, Foundations of Clinical Research; Applications to Practice, Edition 2, page 137-149

[2] Inclusion and Exclusion Criteria | Examples & Definition (scribbr.com)

[3] What Is Probability Sampling? | Types & Examples (scribbr.com)

[4] What Is Non-Probability Sampling? | Types & Examples (scribbr.com)