Introduction to University studies
There seemingly exist merely two ways to teach statistics. I encountered one in my shitty beyond salvation Canadian college, the other at my alma-mater TSNUK. I learned a lot in the latter, but none of that was statistics. I learned barely anything in the former, I did not even know the formula for the normal distribution, yet I learned what statistics is about and how to apply it, and had no trouble continuing mastering it by myself. And that’s what matters in the end, being the difference between the quality education and pretence of it.
Given the actual unsatisfactory amount of milk in each of the 5 packs (say, all 985), and the 1000 ml +- 2% guarantee by the producer, how do you determine whether you are being fooled? You construct the null hypothesis H_0 being “You are not fooled, your sample just sucks a bit” vs the alternative “You are fooled, the producer consistently pours less,” give the producer a 5% chance of error, and then under the assumption of normality of the amount of milk poured into a pack with the known standard deviation 10 ml (if producer means 95% guarantee) you compute the z-score of your sample and look up its p-value, aka the probability to obtain such sample mean or smaller given the null hypothesis. Then, if the p-value is smaller than 5%, aka your allowance for the producer’s error, you go argue with them, having successfully rejected the hypothesis of his honesty. Otherwise, you have failed to reject such hypothesis, and either need more data or to put up with it.
At TSNUK I learned a lot about distributions and transforms, the multidimensional distributions aka vector ones and finally even about the limiting distribution of the standard deviation of the sample. However, I never had a single example to make the use of it, and even if there was a general statistical theory, the way of its teaching was so dry I could have easily omitted it. Again, there was never a single empowering example that would show you what that is about. And as one can see above, one example tells you whole lot more than arbitrary amount of barely applicable math.
I believe that every course should start with such empowering example, and then everything else would stick by itself. The students would themselves question their power and try to extend it further, so that teachers won’t need applying any effort conveying more difficult concepts and more technical details. It will all come together effortlessly once you get a taste of that power.
There seemingly exist merely two ways to teach statistics. I encountered one in my shitty beyond salvation Canadian college, the other at my alma-mater TSNUK. I learned a lot in the latter, but none of that was statistics. I learned barely anything in the former, I did not even know the formula for the normal distribution, yet I learned what statistics is about and how to apply it, and had no trouble continuing mastering it by myself. And that’s what matters in the end, being the difference between the quality education and pretence of it.
Given the actual unsatisfactory amount of milk in each of the 5 packs (say, all 985), and the 1000 ml +- 2% guarantee by the producer, how do you determine whether you are being fooled? You construct the null hypothesis H_0 being “You are not fooled, your sample just sucks a bit” vs the alternative “You are fooled, the producer consistently pours less,” give the producer a 5% chance of error, and then under the assumption of normality of the amount of milk poured into a pack with the known standard deviation 10 ml (if producer means 95% guarantee) you compute the z-score of your sample and look up its p-value, aka the probability to obtain such sample mean or smaller given the null hypothesis. Then, if the p-value is smaller than 5%, aka your allowance for the producer’s error, you go argue with them, having successfully rejected the hypothesis of his honesty. Otherwise, you have failed to reject such hypothesis, and either need more data or to put up with it.
At TSNUK I learned a lot about distributions and transforms, the multidimensional distributions aka vector ones and finally even about the limiting distribution of the standard deviation of the sample. However, I never had a single example to make the use of it, and even if there was a general statistical theory, the way of its teaching was so dry I could have easily omitted it. Again, there was never a single empowering example that would show you what that is about. And as one can see above, one example tells you whole lot more than arbitrary amount of barely applicable math.
I believe that every course should start with such empowering example, and then everything else would stick by itself. The students would themselves question their power and try to extend it further, so that teachers won’t need applying any effort conveying more difficult concepts and more technical details. It will all come together effortlessly once you get a taste of that power.
👍3
Якщо ви задавались питанням, чому в adm:t так мало треків, а останні були випущені в 2022, то ось відповідь. Сподіваюсь, він повернеться цілим і здоровим, як і всі наші Захисники, і продовжить творити.
Підтримайте його треки прослуховуванням аби якісна українська музика жила, а Захисників донатом, аби вони якомога скоріше знищили рештки росні, і народ-гній припинив заважати нам жити і творити.
Підтримайте його треки прослуховуванням аби якісна українська музика жила, а Захисників донатом, аби вони якомога скоріше знищили рештки росні, і народ-гній припинив заважати нам жити і творити.
❤2
Машинний викладач ∆ | #УкрТґ
Stanford prison experiment debunk u mean? Shite, I am literally those guys
I hope you guys love me for the memes I snatch and not my shitëurnalling.
🥰2
Машинний викладач ∆ | #УкрТґ
Stanford prison experiment debunk u mean? Shite, I am literally those guys
Actually every time I hear “Schrödinger’s cat” I wanna throw smth at them.
👍6
Man u don’t wanna know what they turned Olympic fencing into
🌚3