Data Sampling › Population versus Sample {#population-versus-sample}

At the heart of sampling theory lies the distinction between a population and a sample. The population represents the complete set of all individuals, items, or observations about which we want to draw conclusions. This might be all registered voters in a country, every transaction in a database, or all possible measurements from a manufacturing process. The population defines the scope of our inference: the group to which our conclusions should apply.

In contrast, a sample consists of the subset of observations we collect and analyze. When we survey 1,000 voters from a population of millions, conduct quality tests on 100 products from a production run of 10,000, or analyze data from 500 customers out of a customer base of 100,000, we are working with samples. The sample is a window into the population.

Population characteristics are described by parameters, which are fixed but typically unknown values such as the true population mean ($\mu$) or population proportion ($p$). Sample characteristics are described by statistics, which are calculated values such as the sample mean ($\bar{x}$) or sample proportion ($\hat{p}$) that we compute from observed data. A fundamental goal of statistical inference is to use sample statistics to estimate population parameters and quantify our uncertainty about those estimates.

The relationship between samples and populations introduces a key challenge: samples rarely perfectly mirror their parent populations. Random variation means that different samples from the same population will yield different statistics, a phenomenon known as sampling variability. Understanding and accounting for this variability forms the basis of inferential statistics and confidence interval construction.