Today, let's talk about hypothesis testing. This is an ancient method. In recent years, has become popular, which has ushered in a new wave of renaissance in hypothesis testing methods. Many small partners are asking: How to do hypothesis testing? What does that pile of statistical symbols that I don't understand mean? Today, helped you organize a lazy strategy, everyone can copy it and use it. There may be inaccurate descriptions, please don't spray them. 1. What is hypothesis testing? There are two most common scenarios for hypothesis testing: 3 minutes to understand hypothesis testing In short, it is a must for quarrels, and a weapon for scolding people.
When is the right time to mobile number list use hypothesis testing? When the error is obvious, no hypothesis testing is required. For example, the performance of sales and operations is not good, you can directly take the sales performance and operation indicators of the past X days. If the sales performance has been declining, there is nothing to say. At this time, there is no need for hypothesis testing, just open up. Some problems are not so obvious, such as the product life problem mentioned above. Unless the product is used badly, I don't know what the life expectancy is.
When users encounter the problem of short product life, they will only change the brand silently, and rarely complain about "not durable". At this time, sampling testing must be done, and hypothesis testing methods must be used. When demonstrating a new idea is very good, it is also necessary to make the new idea first, and then do a small-scale test. At this time, the hypothesis testing method should also be used. In conclusion, the hypothesis testing method is suitable for sample testing/small-scale testing scenarios. 3. How is hypothesis testing performed? As we all know: falsification is much easier than proof.