What are confidence intervals? Actually.
HTML-код
- Опубликовано: 28 июн 2024
- See all my videos at www.zstatistics.com/
1:01 Intuition
4:46 How are they calculated
14:28 Confidence interval examples
18:50 Frequentist vs Bayesian
ERRATA: The confidence interval created for the population proportion at 13:51 should be 0.03 to 0.21.
HEALTH STATS IQ PLAYLIST
• Health Stats IQ
T-DISTRIBUTION VIDEO
• What is the t-distribu...
BINOMIAL DISTRIBUTION VIDEO
• Binomial Distribution ...
Your content and intuitive approach is seriously life-changing, statistics went from this impossible to understand enigma to something I actually enjoy and dedicate time to learn in my spare time. I would love to see your approach to real world statistical problems (as you alluded to in your previous comment) if you ever get around to it!
Agreements, Super, agreements. I went from watching Zedstats for another explanation/more insights into a public health hypothesis testing course I was taking to watching these videos with my cat on Sundays for fun.
Couldn't agree more. This guy rocks.
Thank you so much, Justin! As a father of 2 working full time and going to college, your videos have made my understanding of statistics go from impossible to manageable.
You are seriously amazing. Can't thank you enough for putting out this content!
I just keep adding your videos to my playlists, amazing work.
Brilliant, I watched many videos to understand what those figures mean, and you are the 1st one that explains it for me.
22:27 Thanks for great explanation of the fundamental misunderstanding of confidence interval. This describes this perfectly!
Thank you so much. I have been swimming in the formulas all semester. I'm three minutes in and when you explain without any maths it helps make so much sense. I need more people to teach maths this way instead of leaving us drowning in the formulas and problems.
Hi Justin,
This course is invaluable, so glad I stumbled upon it this week. I need this content in preparation for a role I will be taking on in the near future.
Thank you so much.
Thank you. I particularly liked how you showed how to interpret published data in text and visual form.
Outstanding. As an instructor, I will be sharing this in class. I expect my students will find more of your material of value as well. Cheers!
As an aspiring data analyst-I found your channel to be inspirational. Thank you
Thank you Justin .. always waiting for your new videos 💪🏻
Thanks so much man, you are a genius. I do love your content, teaching style and explanation. Such a talented teacher.
You just hit the spot with your explanations. super thanks
Excellent, very accessible explanations! Thank you!
great choice of examples and well-explained, thankyou!
I have been following your videos for a long time now, I really like it , you have changed the way I looked at statistics. Now my question, WHY THIS HAIR CUT?!
You are awesome, Zedstatistics. Thankyou for all the lessons, :)
Clever and user friendly. Glad I found you!
This was VERY helpful, thank you, Justin!
Mark my words. One day this channel will blow up and everyone will be like WHY DIDN'T RUclips RECOMMEND THIS TO ME EARLIER!!! WHY!!!!
This channel deserves more than a million subscribers!!... Go ahead sir❤️❤️❤️
Mate, you're a legend!!
Thank you for the presentation!
You make this so understandable.
Personally, I love using elections for confidence interval examples. I also really like your comparison between Frequentists and Bayesian intervals. You don't see that very often.
I just put my own video up about how to interpret CI's after talking with friends and family over the 2020 election.
Thank you for your work, Justin! Watch you videos for the exam prep, but will definitely come back to watch all of your videos for a better conceptual understanding. I have a small question: why do we use t table in this example with women heartrates, if we have 50 observations? I thought that everything over 30 is considered to be of a Z distribution?
Very very clearly and conspicuously explained sir. Thank you.
Nothing but impeccable! just love you! Can you please provide videos on survey sampling techniques (SRSWOR, Stratified, Systematic, etc. and Ratio and Regression methods of estimation) please...
This is very helpful. Thank you.
Excellent videos, really easy to understand.
Thank you so much after how many videos i saw yours is best one big 👍
Great videos, contents and presentation. What software do you use to create the pan and zoom presentation with the colored bubbles with text inside them?
Hi, Very nice work and I am a big fan. Is there any chance that you might give us some enlightenment on endogeneity and panel data (fixed effect vs Random effects) in the near future ? Thanks and continue your great work
thank you very much my friend, that was appreciated :d
Great video, thank you!
Yes man! I'm 27 and have been watching your lessons for months now, wish I'd had your lessons when I went to school! Thanks a lot and keep up the good work, peace from UK! :D
Great video. Thank you
Thanks a lot.
huge thanks
Great video sir. Thank you
i ADORE U SO MUCH THANK U SO MUCH
Great video!
Thanks a lot!!
Thanks!
thank you so much
amazing thank you so much
Hi! This video is great!... I was wondering if the confidence interval of this video is like in the middle of the Frequentist and Bayesian, since theta is fixed, however it will be in the middle of the distribution courve... so... It will be grate if you can clarify this. Thanks a lot!
thank you. ^^
Excelente explicación.
You are the best.
Thank you so much for sharing this video. I have few query regarding the topic whether it is the same thing as bias corrected accelerated confidence interval.
Just a question for clarification: did I get it right, whether I can simply infer from the fact that the error bar crosses the vertical, dotted line, that the difference is not significant compared to the reference category? Is it always the case?
Awesome work.... Thanks for all the videos... What are your plan for this year?? Which topics will you cover?
I guess at some point all the important concepts will be covered. However, I would definitely love some statistical simulations, which makes things way less abstract and develops an intuititive feel to it!
Yo RRR, next videos in the production line:
1. Type I and Type II errors (and Power) - my favourite explanation for this!
2. Frequentist vs Bayesian statistics - referenced in this video
3. Series on survival analysis
4. Series on time series
Subject to change :)
Am going to start a new channel dedicated to interesting statistical/mathematical problem questions. This will depend on other work commitments but I've got some good ideas.
Hmmm good point, LV. Will try to include some more simulations! People went wild when I put one in the multicollinearity video - so will do more :)
@@zedstatistics Any chance the Type I and Type II errors will come up before my statistics exam on Monday?
Amazing lecture, I wonder why people make thumb down?
Love your videos, and your style of teaching! Could also please do a series on Calculus?
Would be happy to, however I think it MUST have been done already by many other channels. Will keep it in mind after I complete the videos that are already in the works :)
@@zedstatistics Many thanks! I just love your style of teaching
Calculus, if taken by you, would also be so intuitive and great to learn. please consider this request Justin. Thanks a lot.
You rock!
Too good! 👌🏾❤️x
Sry for the question (especially if i just have not been looking well enough) but are is the excel spread sheet mentioned 6:48 still available?
Hi! I have question I have been trying to look all around and I cannot find anywhere, so I guess I am just terribly missing something but I don't know what. So I get how the calculations of the confidence interval are done, but I still don't understand how we pass from getting the values that capture the 95% of the probability distribution of this sample mean to say: and we know that the true population will be included here 95 per cent of the times, were we to repeat the experiment 100 times. I still don't understand this step, from reconstructing the confidence interval to infere that the true population mean should be within there with a 95% confidence. Could you help here? thanks!!
Hi Justin, if 'mu' is at the center of the sampling distribution why don't you fit a normal distribution to the sample-means and get true-mean of the population?
still waiting for that FREQUENTIST vs BAYESIAN VIDEO
in August.
How can i say no to the muffin man? I'm on it, MM1. (Though might be a few weeks yet. Busy as, right now).
Why are you using the Student-t distribution (9:04)?
The sample has more than 30 observations (n=50).
Thank you for sharing.
EXCELLENT placement of ads, was that your choice? Or are they random
I mean to say, they didn't break my concentration or your flow of ideas
Weird thing to complement you on? Idk but it makes a huge difference
I get so confused when we talk about standard error of a binomial distribution. I know SE = s/(n^(1/2)) and s for binomial distribution is np(1-p). So, how do you get the standard error that you used in the video? It perfectly fits if instead of Binomial distribution, we use Bernoulli's distribution
23:00 Not only hardcore statisticians! the Frequentist CI is a bit of a "desaster" as even Neyman himself was not happy about it. And the fact that it is applied a Bayesian interpretation just shows you how screwed up Frequentist logic is in hypothesis testing and parameter estimation :)
Are confidence intervals used when you're looking at a distribution of one single sample, or when you're looking at a distribution where the values are means of various samples (with the end goal of determining the population mean)?
I
What is the ratio of including music in your video ? should we listen the music or your lecture. Is it statistic learning video or relaxation video?
🏆 👑 🥇
sir here the proportion (p) you've mentioned is same as p values?
What is the intuitive approach ?
what are the D,F ?
3:51 Isn't that Credible Interval, which is different from Confidence Interval?
im from nsw too :D
Why did I think he was holding one of those chickens with the fluffy heads ?
YOU MADE LOVE STAT GREAT APPROACH
11:28
What’s the point of taking a 95% confidence if it covers almost 100% of the range of values?
Might as well take the sample mean that we calculated?
We are 95% sure/ there’s a 95% probability that the actual population mean lies b/w 72.1 & 79.7
Cool haircut
The presenter or the microphone?
That is unclear for me why does in Marcella Alsan research ar 15:32 as a reference value mean is used. Should it be the confidence intereval for mean value and not mean value be used?
Just noticed that you had a corona virus in the background while you gave the introduction.
Nice content! I checked your 95% CI for the population proportion (since you asked for checks in your video) and I didn't find your result.
p = 0.12 ; n = 50 ; z95 = 1.96 => p +/- sqrt(p(1-p)/n) = [0.0299, 0.21]
I'm quite confident in my computations, but I invite anyone to double-check it.
I got the same answer you did. [0.0299,0.21]
why p uses z, while left uses t from the same data?
why we are using T distribution @9.35
when n is 50 shouldn't we use Z statistic.
What kind of dog is that? Just kidding. Very useful video. I;ve subscribed
You are using the sample std for continuous data (and t distribution) and populational std for discrete (Bernoulli) data. Why?
why do we use 95% Confidence Interval?
10:07
Hello Tom ellis.
Good content.
Negative point : the sound is awful.
Short videos are more helpful
Am I imagining or you do use your handsomeness as an advertising method
Hi guys. I´m trying to be a better reader of clinical literature and I encountered this sentences in an medical article that I´m interested in. I would like to understand this appropriately:
Multivariate predictors of poor outcome were persistent neutropenia (hazard ratio, 5.43; 95% confidence interval, 2.64 to 11.11) and recent corticosteroid therapy (hazard ratio, 2.18; 95% confidence interval, 1.98 to 3.96). The actuarial survival of patients was 0% for patients with both unfavorable prognostic factors and 4% for those with persistent neutropenia only. By contrast, patients who had no risk factors or whose only risk factor was corticosteroid therapy had a 67% and 30% survival rates, respectively (P < 0.0001).
What is this Hazard ratio and why is it described next to CI and how should I understand this p(roportion?) of 0.0001. Sorry for not knowing and thanks for your help
hazard ratio is risk at a time - if HR=1 "persitent neuropenia"/"corticosteroid therapy" would not increase the risk of poor outcome, if hr >1 it would show that the poor outcome is increased by it. The CI is necessary as it should not include 1 (because if its 1 or lower it could mean that it actually does not increase the risk). As the CI only includes numbers >1 we now can say that "persistent neuropenia"/"corticosteroid therapy" really increases the risk for poor outcome. This p
uli wamu last wechipondo
Note for future revision.
4:18 20:53
"This is not the internal that will hold the weight of 95% of the apples in the orchard. It is the interval that we are 95% confident will contain the true and known value of the population mean."
"If we were to take a lot of samples of similar size, and create 95% confident intervals from them (e.g. with bootstrapping), 95% of them will contain the true population mean. But 5% of them will not contain the population mean."
It's the interval within which the population mean is likely (95% sure) to lay.
Let's say CI = 50-90, then I'm confident the mean is either 50, 51, 52, 53,... 88, 89 or 90. It could be 51 or 89. If it's 89, then the credible interval can be 89 +/- a value from std dev.
why are you adding background music.. just annoying
What's alarming is that 83% of 'indigenous' people live the main cities and regional towns. Their experiences should be the same as the rest of Australia. So their obesity should be the same.
Only 17% live in remote communities but they should be living healthier lives but are being served processed foods.