Get AI summaries of any video or article — Sign up free
LESSON 28 - SAMPLING: DEFINITION OF TERMS USED IN SAMPLING, PURPOSE & LIMITATIONS OF SAMPLING thumbnail

LESSON 28 - SAMPLING: DEFINITION OF TERMS USED IN SAMPLING, PURPOSE & LIMITATIONS OF SAMPLING

5 min read

Based on RESEARCH METHODS CLASS WITH PROF. LYDIAH WAMBUGU's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Sampling is a strategic decision to study a subset of a population, enabling credible results without collecting data from everyone.

Briefing

Sampling in social science research is a strategic choice to study a subset of a population rather than everyone, making it possible to produce credible findings without collecting data from every member. The core idea is that researchers must define key terms—population, elements, units of analysis and observation, target versus accessible populations, samples, sample size, and the sampling frame—because these concepts determine who gets studied, what can be generalized, and how results are interpreted.

A population is the entire group of people, events, or things of interest that a researcher wants to investigate. Within that population, each individual is an element, and the unit of analysis is the primary unit used in statistical analysis (the “who or what” the study’s questions focus on). The unit of observation is different: it is the object from which information is collected. Researchers typically distinguish between the target population—where findings are intended to generalize—and the accessible population—the portion they can realistically reach. For example, if the target population is all undergraduate students in a Bachelor of Arts program, the accessible population might be only those in the same program at the University of Nairobi.

From the accessible population, researchers draw a sample, which is a subset of the study population. Samples can be representative or exploratory. A representative sample mirrors the larger population’s characteristics and is usually selected randomly so every member has an equal chance of selection; this supports generalization and is commonly linked to quantitative research. An exploratory sample is “information rich,” often used in qualitative research to probe issues and generate new ideas or theories.

Sample size (n) is the number of subjects included in the study, and it is determined after establishing the sampling frame. The sampling frame is a list of all elements in the population from which the sample will be drawn. A good sampling frame is relevant (contains only elements tied to the research problem), complete (covers all relevant elements), precise (excludes irrelevant information), and up to date (kept current so outdated entries don’t distort selection). Randomization is emphasized as well: each individual should have an equal opportunity to be selected.

Several related concepts help interpret results. A parameter is a numerical characteristic of the population (e.g., μ for the population mean), while a statistic is the corresponding numerical characteristic of the sample (e.g., x̄). Precision describes how accurately sample-based estimates reflect population parameters. Response rate measures the percentage of selected participants who actually take part.

Sampling is used for practical reasons: it is more economic, faster, feasible for large populations, necessary when some groups are inaccessible, appropriate when observations are destructive, and—counterintuitively—can be more accurate than a census when conditions make full enumeration unreliable. Still, sampling has limitations. Sampling error introduces uncertainty because those not selected may differ in attitudes, knowledge, or perceptions. Sample bias can occur when the sampling frame is incomplete, imprecise, or outdated. Finally, finding a good sampling frame can itself be difficult, which can undermine the quality of the sample and the credibility of inferences.

Cornell Notes

Sampling in social science research is the deliberate selection of a subset of a population to study, rather than collecting data from everyone. Clear definitions matter: the target population is where results should generalize, the accessible population is what researchers can reach, and the sampling frame is the list used to draw the sample. Representative samples support generalization through random selection, while exploratory samples are “information rich” and help generate new ideas in qualitative work. Sample size (n) is the number of subjects, and concepts like parameters (population values) versus statistics (sample values), precision, and response rate guide interpretation. Sampling is used because it is economical, timely, feasible, and sometimes necessary, but it carries risks like sampling error, sample bias, and difficulty obtaining a strong sampling frame.

How do target population and accessible population differ, and why does that distinction matter?

The target population is the group of individuals with characteristics of interest to the researcher, and to whom findings are intended to generalize. The accessible population is the portion of that target group the researcher can realistically reach. Because sampling is drawn from the accessible population, researchers must ensure it is representative of the target population; otherwise, generalization becomes weaker. Example: if the target population is all Bachelor of Arts undergraduates, the accessible population might be only those in the Bachelor of Arts program at the University of Nairobi, leaving out students in other universities.

What is a sampling frame, and what makes a “good” sampling frame?

A sampling frame is the list of all elements in the population from which the sample will be drawn. It should be relevant (includes elements directly tied to the research problem), complete (covers all relevant elements), precise (excludes irrelevant information), and up to date (cleaned regularly so only current, eligible elements remain). The frame is crucial because it determines who has the opportunity to be selected; a poor frame can lead to bias.

When should researchers use a representative sample versus an exploratory sample?

A representative sample is used when the goal is to generalize findings to the larger population. It matches the population’s characteristics and is typically selected randomly so every member has an equal chance of selection, aligning with quantitative research. An exploratory sample is used when the goal is to probe an issue and discover new ideas or theories; it is “information rich” and is mainly associated with qualitative research.

How do parameters, statistics, and precision relate to each other?

A parameter is a numerical value describing a population characteristic (e.g., μ for the population mean). A statistic is the corresponding numerical value computed from the sample (e.g., x̄ for the sample mean). Precision refers to how accurately the sample-based estimates approximate the population parameters. Since researchers infer population values from sample results, precision is central to judging how trustworthy those inferences are.

What are the three main limitations of sampling, and what causes each?

First, sampling error: people not included in the sample may differ in attitudes, knowledge, or perceptions, creating uncertainty and margins of error. Second, sample bias: it arises when the sampling frame is not relevant/complete/precise/up to date, so the sample systematically misrepresents the population. Third, availability of sampling frames: finding a good, complete sampling frame can be challenging, which can limit sample quality and the credibility of conclusions.

What does response rate measure, and why is it important?

Response rate is the percentage of people selected for the study who actually participate. It matters because non-participation can reduce the effective sample and can introduce nonresponse-related distortions, undermining the representativeness of the sample and the reliability of inferences.

Review Questions

  1. What are the differences between unit of analysis and unit of observation, and how could confusing them affect a study?
  2. Why can a sample be more accurate than a census in some situations, and what practical factors drive that claim?
  3. How do sampling error and sample bias differ, and what specific steps in sampling design help reduce each?

Key Points

  1. 1

    Sampling is a strategic decision to study a subset of a population, enabling credible results without collecting data from everyone.

  2. 2

    Target population is where findings should generalize; accessible population is what researchers can reach, and sampling is drawn from the accessible group.

  3. 3

    A representative sample supports generalization through random selection, while an exploratory sample is information-rich and supports qualitative discovery.

  4. 4

    A sampling frame is the list used to draw the sample; it must be relevant, complete, precise, and up to date to avoid bias.

  5. 5

    Parameters describe population values, statistics describe sample values, and precision reflects how accurately sample estimates mirror population parameters.

  6. 6

    Sampling is used because it is economical, timely, feasible for large or inaccessible populations, sometimes necessary when observations are destructive, and can still yield accuracy.

  7. 7

    Sampling limitations include sampling error, sample bias from weak sampling frames, and the practical difficulty of obtaining a strong sampling frame.

Highlights

Sampling frame quality is a make-or-break factor: relevance, completeness, precision, and currency determine whether the sample can represent the target population.
Representative samples rely on randomization so every member has an equal chance of selection, supporting generalization in quantitative work.
Sampling error and sample bias create different threats to validity—one comes from uncertainty due to who is left out, the other from systematic problems in the sampling frame.
Response rate tracks how many selected participants actually participate, affecting the effective sample used for inference.

Topics

  • Sampling Definitions
  • Target vs Accessible Population
  • Sampling Frame
  • Representative vs Exploratory Samples
  • Sampling Limitations

Mentioned