The census is the best way to collect the most exact information about a population, because in a census, we question every individual in the population. In practice, this method is not always feasible, such as when trying to gather information about crowds, animal populations, or the spread of disease.
To get around the difficulties of a census, we most often proceed with a survey.
There are several methods that can be used to create a sample of a target population.
Each method has advantages and disadvantages.
- Random sample
Every element in the population has the same probability of being part of the sample.
- Systematic sample
Every element is chosen based on a regular rule within the target population.
- Cluster sample
First, we divide the population into clusters (sub-groups of the population) based on, for example, the geographic position of the target population, and then we randomly select a certain number in order to form the sample.
- Stratified sample
Based on a characteristic of the target population, we divide the population into stata (sub-groups of the population) and then randomly select elements from each strata, respecting its proportionality in the population.