samples

Samples are a subset of units drawn from a larger population, used to infer properties of that population or to test hypotheses. A good sample aims to match key characteristics of the whole and to minimize bias. Sampling plans specify what to select, how many units to include, and how to collect data, often under practical constraints such as time and cost.

In statistics and research, common methods include probability sampling—random, stratified, systematic, and cluster sampling—and non-probability sampling, such as convenience samples. Sample size affects precision; larger samples reduce sampling error but may be more costly. Bias, nonresponse, and measurement error can distort results even with careful design.

In science and industry, samples are collected from environments, biological materials, or manufactured products to test quality, safety, or composition. Biological samples may require laboratory processing, chain of custody, and ethical approvals. Environmental sampling tracks pollutants; quality control sampling checks product consistency. Proper handling, labeling, and documentation are essential to ensure traceability.

In music and media, sampling refers to reusing a portion of an existing recording in a new work. Digital sampling raises legal and ethical considerations, including copyright and licensing. In data science and computing, sampling underpins techniques such as bootstrapping and training-test splits, where a subset of data is used to build or evaluate models.