What are confidence intervals? Actually.

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 ...

Пікірлер: 112

  • @v2blaster228
    @v2blaster2284 жыл бұрын

    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!

  • @m.c.degroffdavis9885

    @m.c.degroffdavis9885

    3 жыл бұрын

    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.

  • @thestandarddeviant7932

    @thestandarddeviant7932

    3 жыл бұрын

    Couldn't agree more. This guy rocks.

  • @ryanmills5343
    @ryanmills53433 жыл бұрын

    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.

  • @BrunettePanther
    @BrunettePantherАй бұрын

    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.

  • @leogaussbell1622
    @leogaussbell16223 жыл бұрын

    I just keep adding your videos to my playlists, amazing work.

  • @prathameshmahankal4180
    @prathameshmahankal41803 жыл бұрын

    You are seriously amazing. Can't thank you enough for putting out this content!

  • @weston.weston
    @weston.weston2 жыл бұрын

    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.

  • @ikartikthakur
    @ikartikthakurАй бұрын

    You just hit the spot with your explanations. super thanks

  • @abdelrahmanfarag7958
    @abdelrahmanfarag79582 жыл бұрын

    Brilliant, I watched many videos to understand what those figures mean, and you are the 1st one that explains it for me.

  • @ARS1508
    @ARS15083 жыл бұрын

    Mark my words. One day this channel will blow up and everyone will be like WHY DIDN'T KZread RECOMMEND THIS TO ME EARLIER!!! WHY!!!!

  • @mojopin1122
    @mojopin11224 ай бұрын

    As an aspiring data analyst-I found your channel to be inspirational. Thank you

  • @jesseherbert2585
    @jesseherbert25852 жыл бұрын

    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!

  • @fares5903
    @fares59034 жыл бұрын

    Thank you Justin .. always waiting for your new videos 💪🏻

  • @pillar7420
    @pillar74202 жыл бұрын

    Thank you. I particularly liked how you showed how to interpret published data in text and visual form.

  • @warpiwarpi3533
    @warpiwarpi35333 жыл бұрын

    22:27 Thanks for great explanation of the fundamental misunderstanding of confidence interval. This describes this perfectly!

  • @RohanMishra-mz1gl
    @RohanMishra-mz1gl3 жыл бұрын

    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?!

  • @dsd1610
    @dsd16103 жыл бұрын

    Thanks so much man, you are a genius. I do love your content, teaching style and explanation. Such a talented teacher.

  • @michaelpappas3857
    @michaelpappas3857 Жыл бұрын

    Excellent, very accessible explanations! Thank you!

  • @Popularasian
    @Popularasian3 жыл бұрын

    great choice of examples and well-explained, thankyou!

  • @gharsit_ktm
    @gharsit_ktm3 жыл бұрын

    You are awesome, Zedstatistics. Thankyou for all the lessons, :)

  • @weston.weston
    @weston.weston2 жыл бұрын

    This was VERY helpful, thank you, Justin!

  • @juliecongress6278
    @juliecongress62782 жыл бұрын

    Clever and user friendly. Glad I found you!

  • @jishanzaman3421
    @jishanzaman34213 жыл бұрын

    This channel deserves more than a million subscribers!!... Go ahead sir❤️❤️❤️

  • @Vidobeats
    @Vidobeats3 ай бұрын

    thank you very much my friend, that was appreciated :d

  • @AndrzejFLena
    @AndrzejFLena3 жыл бұрын

    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

  • @onkarkhadke6016
    @onkarkhadke60163 жыл бұрын

    Very very clearly and conspicuously explained sir. Thank you.

  • @swghavoc
    @swghavoc Жыл бұрын

    You make this so understandable.

  • @Streetviews0
    @Streetviews03 жыл бұрын

    Excellent videos, really easy to understand.

  • @sl4896
    @sl48966 ай бұрын

    Thank you for the presentation!

  • @jongcheulkim7284
    @jongcheulkim72843 жыл бұрын

    This is very helpful. Thank you.

  • @rashmi9733
    @rashmi97333 жыл бұрын

    Great video sir. Thank you

  • @just_mudit
    @just_mudit3 жыл бұрын

    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...

  • @dkozlova1
    @dkozlova13 жыл бұрын

    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?

  • @michaelcordova1803
    @michaelcordova180311 ай бұрын

    Excelente explicación.

  • @mahmoudabulmaaty7628
    @mahmoudabulmaaty76283 жыл бұрын

    Thank you so much after how many videos i saw yours is best one big 👍

  • @pedrot
    @pedrot9 ай бұрын

    Great video. Thank you

  • @kristiapamungkas697
    @kristiapamungkas6973 жыл бұрын

    Great video, thank you!

  • @tomquinn7636
    @tomquinn76363 жыл бұрын

    Mate, you're a legend!!

  • @thestandarddeviant7932
    @thestandarddeviant79323 жыл бұрын

    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.

  • @sundargayu2010
    @sundargayu20104 жыл бұрын

    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

  • @shyglo879
    @shyglo8794 жыл бұрын

    Great video!

  • @missh6773
    @missh6773 Жыл бұрын

    amazing thank you so much

  • @leonaphoria6169
    @leonaphoria6169 Жыл бұрын

    i ADORE U SO MUCH THANK U SO MUCH

  • @julianneuer8131
    @julianneuer81313 жыл бұрын

    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?

  • @ophthojooeileyecirclehisha4917
    @ophthojooeileyecirclehisha4917 Жыл бұрын

    thank you so much

  • @abhinavshukla1264
    @abhinavshukla12643 жыл бұрын

    Thanks a lot.

  • @sabiitistephen8278
    @sabiitistephen82782 жыл бұрын

    You are the best.

  • @sandra25
    @sandra252 жыл бұрын

    huge thanks

  • @thebestsnuper
    @thebestsnuper3 жыл бұрын

    Thanks a lot!!

  • @elizabethbass6430
    @elizabethbass64302 жыл бұрын

    Thanks!

  • @lourdesmartinez1486
    @lourdesmartinez14862 жыл бұрын

    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!

  • @teklehaimanotaman3150
    @teklehaimanotaman31503 жыл бұрын

    Amazing lecture, I wonder why people make thumb down?

  • @jongcheulkim7284
    @jongcheulkim72843 жыл бұрын

    thank you. ^^

  • @peerayathongkruer8664
    @peerayathongkruer86642 жыл бұрын

    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.

  • @borisimonovits5364
    @borisimonovits53642 жыл бұрын

    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?

  • @rrrprogram8667
    @rrrprogram86674 жыл бұрын

    Awesome work.... Thanks for all the videos... What are your plan for this year?? Which topics will you cover?

  • @larsvolleberg7897

    @larsvolleberg7897

    4 жыл бұрын

    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!

  • @zedstatistics

    @zedstatistics

    4 жыл бұрын

    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.

  • @zedstatistics

    @zedstatistics

    4 жыл бұрын

    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 :)

  • @larsvolleberg7897

    @larsvolleberg7897

    4 жыл бұрын

    @@zedstatistics Any chance the Type I and Type II errors will come up before my statistics exam on Monday?

  • @skeeterburke
    @skeeterburke Жыл бұрын

    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

  • @manjus3
    @manjus33 жыл бұрын

    Love your videos, and your style of teaching! Could also please do a series on Calculus?

  • @zedstatistics

    @zedstatistics

    3 жыл бұрын

    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 :)

  • @manjus3

    @manjus3

    3 жыл бұрын

    @@zedstatistics Many thanks! I just love your style of teaching

  • @drutsavshukla6131

    @drutsavshukla6131

    3 жыл бұрын

    Calculus, if taken by you, would also be so intuitive and great to learn. please consider this request Justin. Thanks a lot.

  • @alexandermedina4950
    @alexandermedina49502 жыл бұрын

    You rock!

  • @GulzarAhmad-sw1kh
    @GulzarAhmad-sw1kh2 жыл бұрын

    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?

  • @katrindanevova1232
    @katrindanevova12323 жыл бұрын

    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?

  • @damongeo840
    @damongeo8403 жыл бұрын

    Why are you using the Student-t distribution (9:04)? The sample has more than 30 observations (n=50). Thank you for sharing.

  • @Patricia-zi3il
    @Patricia-zi3il3 жыл бұрын

    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!!

  • @deem643
    @deem6432 жыл бұрын

    Too good! 👌🏾❤️x

  • @vivianmac8168
    @vivianmac81683 ай бұрын

    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)?

  • @aviejaypaul6568
    @aviejaypaul6568 Жыл бұрын

    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

  • @zahidkakar7221
    @zahidkakar72212 жыл бұрын

    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?

  • @muffinman1
    @muffinman13 жыл бұрын

    still waiting for that FREQUENTIST vs BAYESIAN VIDEO in August.

  • @zedstatistics

    @zedstatistics

    3 жыл бұрын

    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).

  • @MrAAMNNITAllahabad
    @MrAAMNNITAllahabad3 ай бұрын

    sir here the proportion (p) you've mentioned is same as p values?

  • @oliviaberhan5326
    @oliviaberhan5326Ай бұрын

    Why did I think he was holding one of those chickens with the fluffy heads ?

  • @georgegl3192
    @georgegl31924 жыл бұрын

    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 :)

  • @superfreiheit1
    @superfreiheit12 жыл бұрын

    What is the intuitive approach ?

  • @SabrinaXe
    @SabrinaXe4 ай бұрын

    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

  • @user-vv9lx6oz7o
    @user-vv9lx6oz7o2 жыл бұрын

    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?

  • @karannchew2534
    @karannchew25343 жыл бұрын

    3:51 Isn't that Credible Interval, which is different from Confidence Interval?

  • @kasasaismail7047
    @kasasaismail70473 жыл бұрын

    YOU MADE LOVE STAT GREAT APPROACH

  • @randomcomments6410
    @randomcomments6410 Жыл бұрын

    You are using the sample std for continuous data (and t distribution) and populational std for discrete (Bernoulli) data. Why?

  • @agape13
    @agape133 жыл бұрын

    🏆 👑 🥇

  • @p.z.8355
    @p.z.83553 жыл бұрын

    I

  • @EricMARTIN-celto
    @EricMARTIN-celto8 ай бұрын

    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.

  • @sl4896

    @sl4896

    6 ай бұрын

    I got the same answer you did. [0.0299,0.21]

  • @anuragmuli7476
    @anuragmuli74763 жыл бұрын

    why we are using T distribution @9.35 when n is 50 shouldn't we use Z statistic.

  • @saltsea9499
    @saltsea94993 жыл бұрын

    what are the D,F ?

  • @fatriantobong2097
    @fatriantobong20972 жыл бұрын

    why p uses z, while left uses t from the same data?

  • @elcapitano1633
    @elcapitano16334 жыл бұрын

    Cool haircut

  • @johningham1880

    @johningham1880

    4 жыл бұрын

    The presenter or the microphone?

  • @zenith_journey
    @zenith_journey3 жыл бұрын

    im from nsw too :D

  • @JavierBonillaC
    @JavierBonillaC11 ай бұрын

    What kind of dog is that? Just kidding. Very useful video. I;ve subscribed

  • @ManuelParodiRamos
    @ManuelParodiRamos Жыл бұрын

    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 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

  • @pedroewert143

    @pedroewert143

    Жыл бұрын

    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

  • @ruoranwang8200
    @ruoranwang82003 жыл бұрын

    Just noticed that you had a corona virus in the background while you gave the introduction.

  • @sarahhalo8849
    @sarahhalo88493 жыл бұрын

    Good content. Negative point : the sound is awful.

  • @khalidshaikh7634
    @khalidshaikh7634 Жыл бұрын

    why do we use 95% Confidence Interval?

  • @moza9835

    @moza9835

    Жыл бұрын

    10:07

  • @HimanshuSharma-oy9ss
    @HimanshuSharma-oy9ss3 жыл бұрын

    Hello Tom ellis.

  • @rashadrashed2912
    @rashadrashed29122 жыл бұрын

    Am I imagining or you do use your handsomeness as an advertising method

  • @karannchew2534
    @karannchew25343 жыл бұрын

    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.

  • @allaboutEducations
    @allaboutEducations3 жыл бұрын

    Short videos are more helpful

  • @chilufyachimba
    @chilufyachimba11 ай бұрын

    uli wamu last wechipondo

  • @codewithwaseem6489
    @codewithwaseem6489Ай бұрын

    why are you adding background music.. just annoying

  • @thedarkman4901
    @thedarkman49014 ай бұрын

    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.