Sampling is an important skill to understand when collecting data. Taking a
sample using recognised sampling methods avoids bias and will give a more representative range of results.
Decisions have to be made about where to got to collect the data, which people need to be asked as well as when and what you are going to ask.
A target sample will focus on a particular group of people or trees or shops
etc.
Samples can be taken in different ways:
• A systematic sample takes only s selected sample at set intervals of
numbers, for example, every fifth or sixth person, house or item. Stones might be measured every five meters.
• A stratified sample takes a proportional sample from each group. For
example, when conducting a questionnaire in a village which has 20 new private houses, 40 old private houses and 40 local authority houses, a representative sample from each housing type is needed. By selecting 25% of each group (5, 10 and 10) a proportional sample is collected. This could also be done with tree types, rock types, age groups etc. This applies to you if you are looking at the number of something (litter, houses, parked cars etc) on streets of very different lengths!
• A random sample selects a sample by chance. You may select people
in the street at random or drop a frame down at random to survey plant species.
In addition, a decision must be made about whether your sampling will be of
points, areas or lines. Transects (or line sampling) are a method which involves collecting data along a straight line. Changes in temperature, vegetation type or size of beach material could be sampled by using a transect line from one point to another.
Whatever, method of sampling is used, sufficient numbers must always be
taken to ensure an accurate representation of features or opinions. The more opinions you collect, the more reliable, or significant, your conclusions will be.