On the statistics of statistical thinking

Statistics is a substantial social science, both of the objective laws of social life, has examined the statistical method statistical basis of inheritance and development of the theory of statistical results, adhere to the statistical nature of the social sciences, so close to statistical theory actual statistical work, and social development in the country has been widely used so-called statistical thinking, that is, the actual work in Statistics, statistical theory, applied Research, we must follow the basic philosophy and guiding principle.


First, the Statistics of several common statistical ideas in statistical ideas include: mean thought, thinking of variation, estimated ideas, related ideas, thought fit to test ideas. Statistical thinking is not a natural formation, to go through statistical concepts, statistical sense , statistical concepts phases Statistical thinking is based on changes in human society needs to carry out various statistical practice, Research and general statistical theory, in order to gradually form a system of statistical thinking as an applied Statistics, it is to learn from the new school of mathematical Statistics nutrition, and more widely used mathematical methods, more and more closely linked, but the idea reflected in the statistics on the school, compared with the general, as well as their own special its basic characteristics: (1) statistical thinking emphasis on methodological and application of the uniform; (2) statistical thinking emphasizes the unity of science and art; (3) statistical thinking emphasizes the unity of objectivity and subjectivity; (4) statistical analysis of the ideological emphasis on qualitative and quantitative analysis of the uniform.

1 mean thinking. Mean the study is to be concise and important representatives mean the concept of almost all statistical theory, the basic idea is statistical. Mean thinking also requires the whole problem, but asked to observe the general development tendency to avoid the interference of the individual by chance, it also reflects the overall concept.

2 variation ideas. Statistical Research similar general characteristics of the phenomenon, its premise is that the overall characteristics of each unit there is a difference of statistical methods is to understand the difference in the number of things. Statistics reflect the variation of the more basic concept is the variance, is that "variation" and "average" the concept of average and variation are the same things on the abstract features and macroscopic measurements.

3 estimated ideology is estimated to speculate that the overall sample, is the same type of things getting from here to understand method using estimation methods have a default: the same sample with the overall nature of the sample can be representative of the population but representative sample of Sex accidental factors, the theory of confidence in estimating the degree of measurement is a necessary step to maintain rigorous logic.

4 related ideas. Things are generally linked to the change, some things go hand in hand were often changed or attendant circumstances were present overall in turn, composed of many individual firms, these individual things are interrelated, and we studied things overall homogeneity is based on the formation and thus, between individuals in general, the overall total and the other is always between interrelated.

5. Fitting idea. Fit is the relationship between different types of things in the abstract representation. Any single relationship must rely on other relationships exist, all real things, relationships were very complex, this approach is the law or trend fitting. the results of fitting the model, reflecting the general trend is the trend expression "changes things and relations as reflected in the number of process models and based on this indicates the possibility."
6 test ideas. Statistical methods always inductive, the conclusion is always with a certain degree of probability, based on local characteristics and laws of the promotion can not be completely out to determine the credibility, the inspection process is to use actual data to test samples Some characteristics of the overall number of prior assumptions are credible.


Second, a number of statistical thinking thinking

1 To change the current existence of some incorrect thinking. British biologist, statistician Galton once said: "Statistics has the extraordinary ability to deal with complex issues, when scientific explorers in the process of advancing set in the way of thorns, the only statistics that can help them open a channel. "But that is not so simple, because the reality we face the problem may be much more complex than imagined. In addition, some people think that way more complex, more Science in the actual analysis of the study, like a simple problem complicated, it seems that in order to show its scientific content. In fact, the real science is to simplify complex issues rather than the pursuit of complexity. Associated with this is that some people statistical inference is that only science, describes the statistics is not science, and extend the expansion to only mathematical statistics is a scientific, social and economic statistics is not science this understanding. This understanding is extremely wrong, at least for the ignorance of the socio-economic statistics in Belgium Kettler only to study probability theory mathematician, and focus on the statistics used in human things, trying to create a statistical improvement of society as a tool of economics and demographics in some modern concepts such as GNP, population growth rate, etc., are the legacy of Kettler and his disciples.

2 To continue to develop statistical thinking. Statistics based on inductive reasoning or the logic of inductive thinking as the main way we all know, there are two main methods of logical reasoning: inductive reasoning and deductive reasoning inductive reasoning is based on the observed data (especially incomplete or poor quality information) to generate new knowledge or to validate a hypothesis, namely to the available data based on general characteristics of induction to arrive at a conclusion. inductive reasoning is based on information in the data chance to discover through necessity. deductive reasoning is the ability to deepen understanding of statistics, especially in the necessity to study and understanding based on chance, the great role. Links to free download http://www.hi138. com
3 To deepen the understanding of data analysis and any statistical data analysis Research are inseparable, because it is necessary to obtain statistical conclusions of the study areas, although the form of statistical analysis of the passage with the times changing, but "from the data extract all the information "or" induction and revealed "as the purpose of statistical analysis has not changed. statistical data analysis for the following three reasons: First, based on the same data will arrive at different, even opposite conclusions of the analysis; two is the analysis of the data we are facing is a defect, or sometimes there is not truth; Third, we are sometimes faced with the analysis of data is massive, people can not start, although statistical data analysis has experienced a descriptive data analysis (DDA), inferential data analysis (IDA) and exploratory data analysis (EDA) phases, analysis methods and techniques has been a qualitative leap, but with the increasing demands of human compared to the problems seems to be increasingly The more so, we must deepen the understanding of data analysis, focus "accurate answers to specific questions and get the data from all valid information," the purpose of expanding Research ideas, to continue to carry out data analysis methods and techniques of research. Share on free download http://www.hi138.com