Methods in health research: Probability and non-probability sampling

Background: Sampling is a crucial step in health research that directly affects internal and external validity. Selecting the appropriate sampling technique minimizes bias and enhances population representativeness. Objective: To review the concepts, types, strengths, limitations, and applications of probability and non-probability sampling techniques in health research. Discussion: Probability sampling, such as simple random, systematic, stratified, cluster, and multistage, provides equal selection chances for all population members, enabling precise parameter estimation and robust inferential statistical analysis. It is ideal for national disease prevalence surveys, population-based program evaluations, or large-scale clinical trials. Non-probability sampling, including convenience, purposive, quota, and snowball, is faster, cost-effective, and useful for accessing hidden populations, such as marginalized groups or individuals with sensitive health conditions, although it limits result generalizability. Method selection should consider research objectives, population characteristics, available resources, ethics, and the type of data required. In certain contexts, combining both approaches can leverage quantitative rigor with qualitative depth. Conclusion: No single sampling technique is universally superior. Probability sampling is best suited for large-scale quantitative studies or when precise population estimates are required. Non-probability sampling is appropriate for exploratory studies, hard-to-reach populations, or when resources are limited. Combining both approaches can yield qualitatively rich yet quantitatively valid data, strengthening the evidence base for decision-making in health research.