Current location - Education and Training Encyclopedia - Graduation thesis - Comprehensive Application of Class C: Statistical Fallacy of Demonstration and Evaluation
Comprehensive Application of Class C: Statistical Fallacy of Demonstration and Evaluation
Demonstration and evaluation is an important issue in the course of Comprehensive Application Ability (Class C), which has been examined five times in the current eight exams. Summarizing the test sites of these five exams, we can find that statistical fallacy is a common type of error in the evaluation of class C argumentative essays.

1. despicable fallacy.

Average fallacy refers to the fallacy of misusing the average, that is, giving the nature of the average to the individuals in the population mechanically, so as to draw a general conclusion according to the illusion of the average. Among them, the arithmetic average fallacy is the most common average fallacy, which refers to the wrong argument that improper application of arithmetic average leads to general conclusions based on the illusion of arithmetic average. The characteristic of arithmetic average is to learn from each other's strong points, make up for the small ones with the big ones, and represent a certain overall level of the target population with the final result. For example, the following example:

The average online shopping price of daily necessities in the UK is 12 USD, that in the US is 15 USD, and that in the Philippines is 1 USD. It can be seen that online shopping for daily necessities from the Philippines is cheaper than online shopping from Britain and the United States.

In the above example? Average price? Is the average, can't get the final by the average? Is it cheaper to buy daily necessities online from the Philippines than from Britain and America? Conclusion: because the average does not mean that all goods in the Philippines are necessarily cheaper than those in Britain and the United States.

2. The data is unparalleled.

The data incomparable fallacy refers to the error caused by ignoring the substantive difference between the statistical object and the sample and mechanically comparing the two data. Comparison should have a comparison object and a comparison basis. In other words, to compare, we must have a reasonable * * * frame of reference. Without a * * * frame of reference, we can't compare the two. Also through an example:

QQ has 820 million monthly active users and more than 600 million QQ space active users. It can be seen that Facebook is still the most popular social platform in the world, but the total number of users in QQ and QQ space exceeds Facebook.

In the above example, by? QQ has 820 million monthly active users and more than 600 million QQ space active users? Can't push it out? The total number of users in QQ and QQ space exceeds Facebook? . Because QQ monthly active users and space active users are different concepts, the number of users of the two cannot be simply added up, and there may be some overlap.

3. Independent data.

Independent data is data that is divorced from the basis of comparison. The evidential function of independent data in argumentation is unconvincing, and it must be compared with relevant data to be convincing. A simple example:

There will be a group of salesmen in the dormitory at the beginning of the school season. According to the data, the success rate of girls' dormitory promotion is 65%. Therefore, the promotion of girls' dormitories is relatively successful.

It is concluded that the success rate of sales promotion in girls' dormitories is only 65%, and it is considered that sales promotion in girls' dormitories is more successful. 65% is an isolated data, so it is difficult to draw a conclusion. Because the success rate of male dormitory in the promotion process is not given, no conclusion can be drawn without comparative data.