To be reasonably sure that your observation will apply to the group you are studying, you should select a sample from the population and ensure that the selection is unbiased. The participants are selected at random to ensure that all the individuals in the group get equal chance of being represented in the sample. This removes the sampling bias in the sample.
