This is the most important thing, thousands of miles away, starting from the foundation soil, the most important thing is the bottom. Statistical thinking, statistical methods, here is the acquisition and collation of market research data, followed by the simplest descriptive analysis, followed by commonly used inference analysis, variance analysis, advanced correlation, regression and other multivariate statistical analysis. Only by mastering these principles can we proceed to the next step.

The second step: the software operation is combined with the analysis model for practical application.

The mainstream softwares about data analysis are (from easy to difficult): Excel, SPSS, Stata, R, SAS, etc. Learn how to operate these softwares first, then use the software to process and analyze the data step by step from data cleaning, and finally output the results, test and interpret the data.

Step 3: Select data mining or data analysis directionally.

In fact, data analysis also includes data mining, but it will be subdivided into analysis direction and mining direction in the work, which are different. Data mining also involves many model algorithms, such as association rules, neural network, decision tree, genetic algorithm and visualization technology.

Step 4: Data Analysis Business Application

This step is also the most difficult step to learn. Different industries and enterprises use different analytical methods. The actual work is to solve business problems, so the ability to gain insight into business is very important, and this ability needs to be accumulated bit by bit in the work. Maybe some regression methods will be used in retail business at present, but other mining methods will be used when switching to e-commerce. Although the business is ever-changing, the analysis method is ever-changing, so mastering the technology and using it in any environment depends only on the accumulation of business experience.

Of course, it would be better to get the certificate of data analyst from CDA.