|Module 1: Producing Data
3. Sources of data > 3.2 Generating your own data > 3.2.2. Generating data from observations
vi. Representativeness of a sample to a population
If you are conducting a survey, then you need to be aware of aspects of your surveying method that can affect the representativeness of your samples to a broader population.
Whenever you take a sample it contains information about the population. However, often this information is difficult to interpret. It would be easier if the sample were representative of the population. When using simple random sampling (SRS), you assume that the outcome of a single sample cannot be predicted. In other words, it is random. However, if many samples are taken, you can expect to observe a pattern. Statisticians call this 'long run regularity'. Random events can be described by laws of probability, which allow you to estimate how often an event will occur in the long run.
This page last updated 20 August 2009