You sir, are no normal distribution
HTML-код
- Опубликовано: 1 окт 2024
- Not even close. Lots of this sort of stuff in textbooks as well. WT...H?
(And just in case anybody thinks that I think I'm in that far right tail --- I'm definitely the midwit here.)
For the normal distribution, the ratio of area between 0 and 1 SD from the mean to that of the area between 1 and 2 SD from the mean is about 2.51. With rounding, the figure states it as 34/14, which is about 2.43. The ratio of the actual areas seen in this plot is about 3.75 (50% larger than it should be). Not even close.
jeremy balka never misses 🔥
Ballers
I literally panicked for a sec and ran to check my central limit theorem video to make sure I didn't copy/paste in an incorrect curve. Fortunately it is correct!
Good to hear :)
Oh boy, slightly off topic but this triggered me a little bit. When my previous company does performance review, the higher-ups require the department heads to calibrate the performance ratings so that the outcome would be a bell curve. Manipulating performance review rating to fit into a manufactured bell curve is the dumbest thing ever, and so unfair when people's merit pay is tied to the rating.
Ha, totally. Its unfair and totally stupid. It's just an excuse not to pay people
This is funny 😂
Sigma based probability joke
McCullough Knoll
Can it be parsed 2 normal distributions with std1 and std2? Gaussian mixture model
A fundamental problem is that no matter how we construct a curve to fit the blue shape, the stated percentages are going to be out of whack. The ratio of 34/14 would be (approximately) correct for the green normal curve, and can't possibly hold for anything that fits the blue shape as there is (relatively) too much area in the middle.
Sir, please make video on time series analysis .Your videos are helpful.
I'm glad you've found my videos helpful. I have to say that time series is a little out of my wheelhouse, and likely pretty far down the list of video topics for me. Sorry :)
I wonder if any symmetric bell curve can be generated by adding more than one gaussian curves of similar height, width (SD) and closely placed means.
hi jbstatistics my friend put one of ur videos on for us and i dont get it but im not in university and he is in ur class so thanks for teaching my friend stat one ur the best maybe i will go to university in the future but not math i dont math
Thanks for the comment! Maybe you guys should make my videos part of your regular party routine :)
All Normal curves are Gaussian; but all Gaussian curves need not be Normal...
Yet the value 34/14 just keeps on keepin' on, unchanging as the hands of time keep ticking. If they didn't put the percentages in, I'd have had no complaint.
U should turn these things into ytb shorts, it's profitable without ads.
I don't really know the deal with shorts, though I looked very briefly when making this one because, well, it was short. The required vertical orientation bugged me a little, in this specific scenario. I might throw others up as shorts, if I think of something that works.
I dont understand...please explain
Implied by the premise, basic shape, and percentages given is that the distribution in question is a normal distribution. But the thing plotted in blue is not a normal distribution, and actually not all that close IMO. The superimposed green curve is normal, and the shape is quite different. I see this a lot, not just in fun internet memes but in publisher texts where they are allegedly illustrating the normal distribution (or other probability distributions). I'm really not sure why.
@@jbstatistics thank you. Because at first glance it did look normal to me
@@jbstatisticsgreat commentary. For the publishers who made this mistake, what would be a better way to talk about the blue distribution?
@@BadEnoughDudeRescues I'm thinking about it in the other direction. Some people refer to the normal distribution, and yet plot something that is not the normal distribution. Some statistics textbooks, in their section on the normal distribution, plot something that is allegedly a normal distribution but in reality is quite different (to the point that I can tell visually right away it is not the normal curve). I don't think that makes any sense in the modern era. When discussing the normal distribution, I would think they would simply plot the normal distribution. The blue curve here is just somebody putting a curve together in illustrator or whatever, but it's not the normal curve. It's like saying, "Here is the value of pi: 3.35". There's nothing wrong with 3.35, it's a fine number. It's just not pi. Same idea that whatever that blue thingy is.