Statistical inference mainly includes statistics and inference.

Statistical inference mainly includes statistics and inference.

Statistical inference is two important aspects in statistics, namely statistics and inference. Statistics refers to the process of collecting, sorting, analyzing and interpreting data, while inference refers to drawing conclusions about unknown quantities or phenomena based on known data and background information.

statistics

Statistics is a scientific method of collecting, sorting, analyzing and interpreting data. It includes many different techniques, such as descriptive statistics, inferential statistics and multivariate statistics. In practice, statistics are often used in forecasting, decision-making, risk assessment and other fields.

The core of statistics is data. The data can be quantitative, such as sales volume, temperature, etc. , or qualitative, such as gender, occupation, etc. The types and sources of data are also varied, such as survey data, experimental data and observation data. In statistics, the quality and representativeness of data are very important to the analysis results.

suggestion

Reasoning is a reasoning process, which draws unknown conclusions based on known information. In statistics, inference is usually based on sample data, and conclusions about population parameters are drawn by using probability theory and statistical methods.

Reasoning mainly includes hypothesis testing and Bayesian reasoning. Hypothesis test is to analyze the sample data, make assumptions about the overall parameters, and test the correctness of the assumptions by probability theory. Bayesian inference is based on prior information and newly obtained data to obtain the posterior probability distribution of the population.

Application case

Let's give an application case to illustrate the application of statistical inference in practice.

Suppose a company wants to test the sales of a new product and decides to conduct market research in 10 cities. After collecting and sorting out the survey data, the company found that this 10 city has a high acceptance of new products. However, the company's goal is to determine whether this new product will be popular nationwide. This requires the use of statistical parameter estimation methods.

Parameter estimation is a method to estimate the overall parameters based on sample data. In this case, the overall parameter is the national acceptance of the new product. The sample data is the result of market research in 10 cities. Companies can use these data to estimate the overall parameters and obtain interval estimates of the popularity of new products nationwide. The interval estimate can be a point estimate plus or minus the confidence interval.