Differences and relations among simple random sampling, stratified sampling and systematic sampling

First of all, the difference

1, different concepts.

Simple random sampling, also known as simple random sampling, pure random sampling and SRS sampling, refers to a sampling method that randomly selects n units from the total number of n units as samples, so that the probability of each possible sample being selected is equal.

Stratified sampling method is also called type sampling method. It is a method of randomly selecting samples (individuals) of different layers from the whole that can be divided into different subgroups (or layers) according to the specified proportion.

Systematic sampling method, also known as equidistant sampling method or mechanical sampling method, takes samples from the population according to a certain sampling distance. To extract samples with a capacity of n from a population with a capacity of n, we can divide the population into several balanced parts, and then extract an individual from each part according to the pre-specified rules to obtain the sampling method of the required samples.

2. Different characteristics

The characteristic of simple random sampling is that the probability of each sample unit is equal, and each unit of the sample is completely independent, and there is no certain correlation and exclusion between them.

Compared with simple random sampling, stratified sampling has more significant potential statistical effects. That is to say, if two samples are taken from the same population, one is a stratified sample and the other is a simple random sample, then the error of the stratified sample is smaller. On the other hand, if the goal is to obtain a certain sampling error level, then smaller stratified samples will achieve this goal.

Systematic sampling is characterized by simple operation and not easy to make mistakes in execution, so people are happy to use it in production site.

Second, contact

The possibility of each individual being drawn in the sampling process of the three sampling methods is equal. The initial part of stratified sampling adopts simple random sampling, and each layer of systematic sampling adopts simple random sampling or systematic sampling.

Extended data:

sampling method

The most basic sampling method of simple random sampling. Divided into repeated sampling and non-repeated sampling. In repeated sampling, the units extracted each time still return to the population, and the units in the sample may be extracted several times. In non-repeated sampling, the extracted units will not be put back into the population, and the units in the sample can only be extracted once. Non-repeated sampling was used in social survey.

The specific methods of simple random sampling are:

Direct selection method

Direct sampling method, that is, direct random sampling from the population. For example, a number of goods are randomly selected from the shelves for inspection; Choose several stalls from the farmer's market stalls for investigation or visit.

draw lots

Firstly, all individuals in the population are numbered (the numbering range can be 1 to n), and the numbers are written on numbers with the same shape and size, which can be balls, cards, paper strips, etc. Then put these numbers in the same box and mix them evenly.

When drawing lots, every time 1 number is drawn, a sample with a capacity of is obtained. When numbering individuals, you can also use the existing numbers. For example, when taking samples from the whole class, you can use the student's student number and seat number. The lottery method is simple and easy, and it is suitable for the situation that there are not many individuals in the whole population.

Random number table method

Random number table method, that is, using random number table as a tool for sampling. Random number table (see example), also known as random number table, randomly arranges 10 numbers from 0 to 9 into a table for future reference. Its characteristic is that no matter horizontal reading, vertical reading or interlaced reading is irregular. Therefore, using this table for sampling can ensure the realization of the random principle and simplify the sampling work.

The steps are as follows:

1, determine the overall scope and arrange the unit numbers;

2. Determine the sample size;

3. Sampling the sample unit, that is, starting from any number in the random number table, reading it in a certain order (up, down, left and right) or at intervals, selecting numbers within the range of numbers, not selecting numbers outside the range, and not selecting repeated numbers until the predetermined sample size is reached;

4. Arrange the selected numbers and list the corresponding cell names.

Examples are given to illustrate how to use the random number table to extract samples.

When the number of initial readings is randomly selected, the reading direction can be right, left, up, down, etc.

In the process of reading every two digits above, a series of two digits are obtained. After removing the unqualified and repeated numbers, the numbers that appear in turn can be regarded as the number of individuals extracted from the population in turn.

Since the probability of which number appears in each position in the random number table is equal, the probability of which two-digit number is read every time is also equal, that is, which individual number is extracted from the population. So the random number table is used to extract samples, which ensures that the probability of each individual being extracted is equal.

Because the systematic sampling method is simple in operation and not easy to make mistakes in execution, people are happy to use it in the production site. If a product is regularly sampled for inspection in a certain process, it can be regarded as an example of systematic sampling method.

Steps:

1, numbering: number the n individuals in the crowd first, and sometimes you can directly use the numbers brought by your own individuals, such as student numbers and house numbers.

2. Segmentation: determine the segmentation interval k and the segmentation number. When N/n(n is the sample size) is an integer, take k = n/n.

3. Determine the first number of individuals: In the first paragraph, determine the first number of individuals l(l≤k) by simple random sampling.

4. Sampling: sampling is carried out according to certain rules. Usually, the second individual number (l+k) is obtained by adding L to the interval K, and the third individual number (l+2k) is obtained by adding K, and so on until the whole sample is obtained.

Baidu Encyclopedia-Simple Random Sampling

Baidu encyclopedia-systematic sampling

Baidu encyclopedia-stratified sampling