18.2.3—Importance of random sampling

Random sampling chooses sampling points by chance so the investigator does not deliberately favour particular parts of the study area. It is most suitable when the area is reasonably uniform and has no clear distribution pattern.

1. Define the whole study area and the population or community to be sampled.
2. Generate random coordinates or point locations across that area.
3. Place the same sampling unit, such as a quadrat where appropriate, at each selected point.
4. Use the same sampling rules and record the observations from every point.
5. Repeat across enough independent points, then judge whether the sample represents the whole area.

Because the locations are selected by chance, random sampling reduces selection bias from the person carrying out the investigation. Standardising the sampling unit and recording rule makes comparisons between points fairer; repeated points improve confidence in the estimate.

Systematic sampling places points by a planned pattern chosen by the investigator, so the chosen pattern or starting point can miss parts of the area or introduce bias. Do not treat random sampling as automatically best for a strongly patterned or clearly non-uniform area.

Random sampling is a method for choosing representative sample locations. It does not by itself prove that a sample is large enough or replace later analysis of abundance, distribution, correlation or Simpson’s index.