A sample needs to be representative of its population so you need to avoid SAMPLING BIAS.
A BIASED SAMPLE IS ONE WHICH DOESNT FAIRLY REPRESENT THE POPULATION.
To avoid sampling bias:
1. Select from the correct population and make sure none of the population is excluded- this means drawing up an accurate sampling frame and sticking to it.
2. Select your sample at random:
- Every person or item in the population has an equal chance of being in the sample.
- Each selection is independent of every other selection.
- To get a truly random sample, you need a complete list of the population- an accurate sampling frame.
- To choose the sample, give every sampling unit a number. Then generate a list of random numbers (using a random number generator) and match them to the sampling units to select your sample.
3. Make sure all your sample members respond- otherwise the results could be biased.
Advantages and Disadvantages of Sampling
Sampling - Advantages
- Quicker and cheaper than a census, and easier to get hold of all the required information.
- Its the only option when surveyed items are used up or damaged.
Sampling - Disadvantages
- Samples can be easily affected by sampling bias.
- Variability betwen samples- each possible sample will give different results, so you could just happen to select one which doesnt accurately reflect the population.
(To reduce variability use a large sample size. The more sampling units that are surveyed, the more reliable the information should be).