Critical thinking and pandemics x: generalization

In a pandemic, similar to Covid-19, could be asked how the number of cases determined and the mortality of a disease is like. Some may be worrying or skeptical because the numbers often change and usually vary in countries, age groups, etnieties and economic classes. This essay provides a basic general view from the samples of the samples of the basic method of inferences from all populations, what philosophers call inductive generalization.

Inductive dominance is an inductive argument. In philosophy, the argument consists of premises and a conclusion. The local is arguing the reasons or evidence that offered to support the effects, which is the claim. Philosophers often distribute arguments induction and deductive. Philosophy is deductible argument, full support to complete the full support provided by the premises. It is an inductive argument that provides (or allegedly) the local support (but less than complete support) for consequences. If the premises of an inductive argument supports the conclusion properly (or better) is a strong argument. If the premises are true, the conclusion is true. If a strong indignor argument has all true premises, sometimes it is called cogent.

An inductive logic is a feature that a strong inductive argument may have a false consequence when all premises are true. This is what is known as an inductive leap: The conclusion always goes beyond the premises. This can also be drawn a conclusion of what has not been seen. David Hume Humen 1700. It said that this could never be sure that inductive reasons and later philosophers called the induction problem. It means practical terms, although using premises that use the perfect inductive reason, our effects can be false. But induction is often the only option, and because we use it because we use it. So when the initial numbers of Covid-19 were wrong, that’s what we expect. The same is expected in the next pandemic.

So then is it inductive generalization? Approximately is an argument based on the consequences of a whole population. The formal version is as follows:

1. Premise: They are observed x% P% y%.

Conclusion: All XS are P% ys.

It would be an observed XS sample and all XS would be the target population. As an example, if someone wanted to know the mortality rate for men over sixty, the target population would be more than seventy years old.

While the argument is easy, a powerful generalization can be challenged when sorting. Without inclusion in statistics and methods for strict generalizations, I will go over the basic assessment method, make sense, when experts talk about such issues in the next pandemic.

We can affect the presence or absence of presence or absence in the sample, with the presence or lack of property concerned, therefore, a sample representative will have these factors in proportion to the target population. For example, if we want to determine the rate of infection for all people, we should make sure that our samples should be all the factors that cause infection rates and our samples should reflect our target population ages, ethnicity, basic health and other important features. The sorting of what important factors that are important can be challenging, especially when it comes to spreading a pandemic. The sample would be representative to the level that is properly reflected in the target population.

A sample is a factor that the factor is not in the sample as far as the same proportion of the population is present. This type of bias sample was a problem while trying to generalize around Covid-19. An example of this was trying to draw a conclusion about the Covid-19 Moneatability. To do this, math is simple (simple calculation of the infected percentage), getting well numbers is hard to get people infected and we needed to know how much they died.

Experts tried to determine the number of polluted tests and modeling, they are also inductive reasoning. In the United States, most of the tests were the people who showed symptoms and this created a naughty sample: not to get a sample sample, even if it should be tested with symptoms. It was also a practical issue and the cause of the issue of the test accuracy and the cause of death. This will also be true in the next pandemic.

To use concrete, but composed of examples, 5% of those who have tested the positive Covid-19 were finished dying, the generalization of this sample would be as strong as a sample representation. If only the patient has been tested, the sample will not be representative and the conclusion of virus ledity will be wrong.

The virus is also a challenge that affects different populations. While there is a general rate of infection and lethality rate for the total population, there are different infection rate and lethality rate for different internal groups of man’s population. As an example, the elderly were more than younger people.

In addition to the representation, the sample size is important; The greater the better. This provides us with two other concepts: a margin of bugs and confidence. The margin of error is a range of points that fall into the result of general generalization; This number is presented according to being plus or minus. The margin of error depends on the level of confidence in the sample size and argument. The level of confidence is usually presented as a number and indicates the percentage of the arguments in the case with the actual conclusion.

Approximately (1o, 000 +) populations, a sample must be 1,000 people as a representative (assuming that the sample is properly taken). This table in Moore & Parker Critical thinking The text shows the link between the sample size and the margin of the mistake (95% confidence level:

It is practical to carry that the size of the sample is important: a small sample will have a large margin of error that can be useless. For example, suppose that a group of 50 Covid-19 patients were fully recovered by hydroxicocloric tablets and 10 of them. Aside from the causal reason (it would be a terrible mistake) The best we can say is that 20 + / – 14% of patients treated with hydroxikocally will be fully recovered. It’s just a simple generalization and a controlled experiment or an analysis should be properly evaluating a causal claim.

There are several missions that can occur with a generalization (errors). I will discuss the next attempt. Be safe and see you in the future.