Top 4 Ways to Learn Statistics

4 Ways to Learn Statistics

Numerous people will say: in school countless times on the statistics course, learn a whole lot of theoretical knowledge, also remembered a lot of formulae, but once encountered the actual data, always don't know how to start, and ultimately can only look for a few articles similar to the above method, regardless of whether it is right or not. The first step is to make sure that you have a good understanding of how to use the data.

Why we can't learn statistics, I guess many people rarely go like this problem. Although I have not intentionally thought about it, but in the years of assisting others to design, analyze the experience, slowly found some possible reasons.

1, statistics itself is a very flexible discipline that really meets the ? Specific problem specific analysis? Such a philosophical concept. Even for the same indicators, the methods used on different occasions may be different. A simple analysis of variance (ANOVA), for example, requires different analytical methods depending on the type of design. Even for the same batch of data, if the purpose is different, the methods and results are also different.

2, the current medical statistics textbooks are mostly a model, cold framework, no humane words, all some people can not understand the words. In fact, this is exactly the sadness of the current field of medical statistics, there are few people who really have the level, most of them seem to understand, lack of experience in analyzing, writing books can only be copied, copying each other, ultimately leading to all the medical statistics textbooks are a face.

3, most people's biggest headache should be: for a batch of data, do not know exactly what method should be selected. Although a variety of methods to apply the prerequisites memorized a bunch, but it seems that each one is like, each and every one is not like. It feels as if it is fine to use any method, just do not know which method is more accurate. There is really no good solution to such a problem. Just like the medical students just graduated from the inability to judge the disease, can only rely on the accumulation of experience.

4. At present, it is a period of irritation in the academic world, and there are not many people who can endure the loneliness to really study the theory. Most people just learn a half-knowledge, and then self-appointed as? Experts? The majority of people just learn a half-understanding, and they are self-appointed as experts. Talent? The hat is very big, not much learning. There are also a lot of people who can learn the theory, but then fail to combine it with practice, detached from the actual problem, with no real benefit. Statistics is a methodology, itself is constantly developing, really want to master statistics, must constantly learn new knowledge, at the same time should be constantly applied, only in the application process can really learn and understand. At present, there are fewer and fewer real masters in the field of medical statistics in China, mainly because most people have difficulty in sinking their teeth into scrutinizing the progress and application of statistical methods. If you don't understand statistics yourself, how can you talk about teaching your students, and the result will be to make them even more confused.

A true master of statistics should at least be familiar with the traditional theory of statistics, understand the latest advances in statistics, often apply statistics to solve a variety of problems, at least proficient in a database tool, at least master a programming language, you must be proficient in SAS, not to mention SPSS, because only in the process of SAS programming, to be able to understand more about the theory of statistics, while SPSS Run only by the menu, in addition to knowing the results, in addition to how to arrive at the results remain ignorant, does not help the understanding of statistics. It is not for nothing that more than 90% of the Fortune 500 analyze their data with SAS rather than SPSS.