The concept of “statistics” means both specific calculated values (averages, medians, etc.) and science, which forms “a set of methods for making decisions in conditions of uncertainty”.
Let's start with the first one: you can't “believe” the numbers. We don't know how many attempts to get the desired result were made by the authors of the article. With a statistical confidence of 95%, any conclusion can be justified with an average of 20 attempts.
But important statistical conclusions can be useful as a hypothesis: you need to take independent data and use statistical techniques to verify their validity yourself. Based on this, you can either verify the conclusions or reduce the degree of trust in the source.
Another problem is that the metric doesn't match the output based on it. Example: in a well-known quote, Zaccaria draws a conclusion about the moral qualities of passengers of different classes based on the metric “the proportion of women and children among the rescued passengers”. But if you use the appropriate metric “proportion of survivors among women and children” and compare it with the proportion of male survivors, the figures are quite comparable: for example, in the third grade, 84% of men (528 out of 706) and 58% of women and children died.
The third problem is the “other things being equal” assumption, which is rarely fulfilled in real life. In the same example with the Titanic, the quote: “in first class … special attention was paid to single women and women with children … Third-class passengers … they only reported the need to go up, and they wandered for a long time through the long corridors of the lower decks, trying to find a way out.”
But mathematical statistics itself (based on probability theory) is a rigorous mathematical discipline that can and should be used. For complex practical problems, it is also worth studying econometrics.
Moreover, without at least a basic level of training in these disciplines, it is extremely difficult to make informed decisions in the modern world.
There is a theory of probability, which, in turn, forms mathematical statistics.
Almost any event can be calculated using statistical data and theoretical verbs.
If you want to read more about how it all works, there is a lot of literature on this subject on the Internet. Because even a popular and clear explanation of the work and principle is quite problematic.
Statistics can be trusted if you know exactly who was interviewed, when, and how. It is clear that if the number of respondents exceeds 10, it is extremely difficult or almost impossible to track such information. But in short, I wouldn't believe it.� My hair is 5% silkier, yeah 🙂 Oh, and maybe “club of leading dog breeders” ask? 🙂
The concept of “statistics” means both specific calculated values (averages, medians, etc.) and science, which forms “a set of methods for making decisions in conditions of uncertainty”.
Let's start with the first one: you can't “believe” the numbers. We don't know how many attempts to get the desired result were made by the authors of the article. With a statistical confidence of 95%, any conclusion can be justified with an average of 20 attempts.
But important statistical conclusions can be useful as a hypothesis: you need to take independent data and use statistical techniques to verify their validity yourself. Based on this, you can either verify the conclusions or reduce the degree of trust in the source.
Another problem is that the metric doesn't match the output based on it. Example: in a well-known quote, Zaccaria draws a conclusion about the moral qualities of passengers of different classes based on the metric “the proportion of women and children among the rescued passengers”. But if you use the appropriate metric “proportion of survivors among women and children” and compare it with the proportion of male survivors, the figures are quite comparable: for example, in the third grade, 84% of men (528 out of 706) and 58% of women and children died.
The third problem is the “other things being equal” assumption, which is rarely fulfilled in real life. In the same example with the Titanic, the quote: “in first class … special attention was paid to single women and women with children … Third-class passengers … they only reported the need to go up, and they wandered for a long time through the long corridors of the lower decks, trying to find a way out.”
But mathematical statistics itself (based on probability theory) is a rigorous mathematical discipline that can and should be used. For complex practical problems, it is also worth studying econometrics.
Moreover, without at least a basic level of training in these disciplines, it is extremely difficult to make informed decisions in the modern world.
Naturally. To the fullest extent.
There is a theory of probability, which, in turn, forms mathematical statistics.
Almost any event can be calculated using statistical data and theoretical verbs.
If you want to read more about how it all works, there is a lot of literature on this subject on the Internet. Because even a popular and clear explanation of the work and principle is quite problematic.
Statistics can be trusted if you know exactly who was interviewed, when, and how. It is clear that if the number of respondents exceeds 10, it is extremely difficult or almost impossible to track such information.
But in short, I wouldn't believe it.�
My hair is 5% silkier, yeah 🙂
Oh, and maybe “club of leading dog breeders” ask? 🙂