Lesson 34: Sampling

Definition

In research, sampling is a word that refers to your method or process of selecting respondents or people to answer questions meant to yield data for a research study. The chosen ones constitute the sample through which you will derive facts and evidence to support the claims or conclusions propounded by your research problem. The bigger group from where you choose the sample is called population, and sampling frame is the term used to mean the list of the members of such population from where you will get the sample. (Paris 2013)

Probability Sampling or Unbiased Sampling

Probability sampling involves all members listed in the sampling frame representing a certain population focused on by your study. An equal chance of participation in the sampling or selection process is given to every member listed in the sampling frame. By means of this unbiased sampling, you are able to obtain a sample that is capable of representing the population under study or of showing strong similarities in characteristics with the members of the population.

Types of Probability Sampling

1. Simple Random Sampling

Simple random sampling is the best type of probability sampling through which you can choose sample from a population. Using a pure-chance selection, you assure every member the same opportunity to be in the sample. Here, the only basis of including or excluding a member is by chance or opportunity, not by any occurrence accounted for by cause-effect relationships. Simple random sampling happens through any of these two methods: (Burns 2012)

1.      Have a list of all members of the population; write each name on a card, and choose cards through a pure-chance selection.

2.      Have a list of all members; give a number to member and then use randomized or unordered numbers in selecting names from the list.

2. Systematic Sampling

For this kind of probability sampling, chance and system are the ones to determine who should compose the sample. For instance, if you want to have a sample of 150, you may select a set of numbers like 1 to 15, and out of a list of 1,500 students, take every 15th name on the list until you complete the total number of respondents to constitute your sample.

3. Stratified Sampling

The group comprising the sample is chosen in a way that such group is liable to subdivision during the data analysis stage. A study needing groupby- group analysis finds stratified sampling the right probability sampling to use.

4. Cluster Sampling

This is a probability sampling that makes you isolate a set of persons instead of individual members to serve as sample members. For example, if you want to have a sample of 120 out of 1,000 students, you can randomly select three sections with 40 students each to constitute the sample.

Non-Probability Sampling

Non-probability sampling disregards random selection of subjects. The subjects are chosen based on their availability or the purpose of the study, and in some cases, on the sole discretion of the researcher. This is not a scientific way of selecting respondents. Neither does it offer a valid or an objective way of detecting sampling errors. (Edmond 2013)

Types of Non-Probability Sampling

1. Quota Sampling

You resort to quota sampling when you think you know the characteristics of the target population very well. In this case, you tend to choose sample members possessing or indicating the characteristics of the target population. Using a quota or a specific set of persons whom you believe to have the characteristics of the target population involved in the study is your way of showing that the sample you have chosen closely represents the target population as regards such characteristics.

2. Voluntary Sampling

Since the subjects you expect to participate in the sample selection are the ones volunteering to constitute the sample, there is no need for you to do any selection process.

3. Purposive or Judgmental Sampling

You choose people whom you are sure could correspond to the objectives of your study, like selecting those with rich experience or interest in your study.

4. Availability Sampling

The willingness of a person as your subject to interact with you counts a lot in this non-probability sampling method. If during the data-collection time, you encounter people walking on a school campus, along corridors, and along the park or employees lining up at an office, and these people show willingness to respond to your questions, then you automatically consider them as your respondents.

5. Snowball Sampling

Similar to snow expanding widely or rolling rapidly, this sampling method does not give a specific set of samples. This is true for a study involving unspecified group of people. Dealing with varied groups of people such as street children, mendicants, drug dependents, call center workers, informal settlers, street vendors, and the like is possible in this kind of non-probability sampling. Free to obtain data from any group just like snow freely expanding and accumulating at a certain place, you tend to increase the number of people you want to form the sample of your study. (Harding 2013)