A range of traditional data sampling methods are available depending on what information (if any) is available for the population. In other words, having some information about the population will assist in determining an appropriate sampling technique. The following need to be considered:

1. Is there information about the population concerning how it is structured (eg. soils may be sandy, loamy, saline), how it is distributed (eg. mostly saline, mostly near a water course) and what is the range of values?
2. Are there specific identifiable groups within the population? For example, in a human population there may be groupings of: children, youth, young adults, middle class and elderly.

There are, then, basically two approaches that can be taken to determining a sampling method:

1. Use minimal information about the population to choose the simplest and most inexpensive sampling technique that is convenient and for which samples are easy to access. In such a case, it is usually not known how well (or not) the sample represents the population. The techniques often used here are referred to as non-probability based sampling methods.
2. Use as much prior information as possible so that the samples are as independent from each other as possible and are representative of the scale and scope of the population. The techniques used here are referred to as probability based sampling methods.

What next?

We will first discuss some of the traditional sampling approaches before proceeding on to the spatial sampling techniques. Note that the spatial techniques have, as their basis, many of the traditional sampling methods.

For you to do...

Suppose that you want to find out if there is support for a proposed recreational complex.

From the description provided here, how would you conduct the sampling using a non-probability based approach versus a probability based approach?