During sampling of granular materials (whether airborne, suspended in liquid, aerosol, or aggregated), correct sampling is defined in Gy's sampling theory as a sampling scenario in which all particles in a population have the same probability of ending up in the sample.
[1] The concentration of the property of interest in a sample can be a biased estimate for the concentration of the property of interest in the population from which the sample is drawn.
Although generally non-zero, for correct sampling this bias is thought to be negligible.
[1] This statistics-related article is a stub.
You can help Wikipedia by expanding it.