1, simple random sampling method:
Its method is the simplest, and there is no distinction or restriction on the whole survey object, which ensures that every survey object has an equal opportunity to become the selected survey object. Basic operation steps: all respondents line up and number.
A certain number of respondents are drawn by lottery (including machine lottery or dice). Conduct an actual survey of the selected survey objects. This method can be used when the total number of respondents is not very large and the individual differences of respondents are small.
2, stratified random sampling method:
The respondents were grouped (stratified) according to different characteristics, and then a certain number of address samples were extracted from each layer by random method. There are two principles in using this sampling method: when layering, there should be obvious differences between layers as far as possible, and each individual in each layer should be flexible.
This can ensure that the samples extracted from each layer can accurately represent that layer. In order to improve the representativeness of the sample, this method can be used to conduct a sampling survey when there is a big difference between each individual in the survey population.
3. Grouping random sampling method:
Divide similar parts of the population into several groups, and then conduct random sampling survey in one or two groups. In practice, when there are too many individuals in the survey population and their positions are scattered, the simple random sampling method can effectively reduce this difficulty.
For example, to investigate the family income of workers in a city, we can divide the city into several districts by random sampling, then divide one district into several streets, and concentrate the investigation on two streets in a certain district, so that the cost and time of investigation will be greatly reduced. When the individuals in the survey are quite different and can only be grouped by region, the method of grouping random sampling can be adopted.
4. Cluster sampling:
Cluster sampling, also known as "cluster sampling", is to divide the whole group into several groups (groups) with small internal differences according to their own characteristics, then randomly select several groups as samples, and then select all the individuals in the selected groups to form samples. The sample has good representativeness and can maintain the integrity of the whole.