OHHH MY GOD THE BEST VIDEO IN INTERNET THAT ACUALLY EXPLAINED P VALULE WITH REALLY SIMPLE EXAMPLES, I WOULD RECOMEND THIS TO ANYONE IF SOMEONE ASK ME! THIS VIDEO IS PURE GOLD! THANK YOU FOR POSTING THINGS THAT MAKE THE DIFFERENCE
Hi from Italy! Finally I can write this very comment! I'VE PASSED MY STATISTICS EXAM :D This is basically thanks to your very clear explanations! The way you break down concepts is just the best! Thank you very much for helping students all over the world.
This is a very late comment, but I case Zed still gets notifications: I’ve struggled with stats since college (40 years ago, ahem), but I’ve been trying to work on my understanding ever since. This vid has finally made p-values understandable. Thank you! Big fan.
Hi Justin! Just wanted to say I was in one of your tutorials for BUSS1020 in my first year of uni and you were honestly one of the best tutors I’ve ever had. You were always so passionate and have made me enjoy statistics so much more! Thanks so much and awesome video!
23:20 - "Eighteen percent probability that the difference we are seeing here occured purely by random chance" Finally. I needed to hear these words. Understanding p-values was quite tough for me, I think i watched 10 video explainations but now it finally clicked in my head. Thank you very much
This is not correct, exactly. I repeated this many times and would like to involve @zedstatistics once and hear his take on it. I believe the correct interpretation would be, "The p-value of 0.18 means that if the null hypothesis (H0) is true, there is an 18% probability of obtaining results as extreme or more extreme than those observed on repeated testing. This is not the same as saying that there is an 18% probability of obtaining the observed effect by chance." Though I understand this is a minor thing since you're rejecting it anyway, for understanding, it's more like percentage vs percentile. It's more about the cumulative probability of extremes vs the probability of this exact finding.
This is amazing. I have ADHD and have always struggled with math in general. Statistics has been absolutely stressful for me during my MPH program. Thanks to you, I feel like I am finally understanding these concepts. THANK YOU!!!
really helpful and clear explanation! I studied it 40 years ago and retrained in a graduate course again several years ago. I still feel unclear with how to interpretate P value. Great contribution!
Hi Zed! I'm Annet from Uganda, I have finally understood both P-value and Confidence Interval after watching your RUclips videos. Thank you so much and Be Blessed.
I have to say this is the best and righteous video about P value, compare to a bunch of high-viewcount videos from blissfully misunderstood author about what p value is. Calculating the p value is not something complex, but the point that you really touch here is: Why rejecting the null hypothesis when the p value reach some small limit (5% to be common)? What is the link of trusting the null hypothesis with getting some really rare samples? And i think you done a great job addressing this stuck point in understanding the p value. Great job. Keep it up!
Hi Zed ! Is it possible that one day you make a serie of videos where you just take a database and examine it juste like you’d do it as a statistician (for example, where do you begin, what questions do you ask yourself and how you answer them etc). I think it would be a perfect example that’ll show us how statisticians work ! Thank you for all your content
This breakdown has thoroughly improved my understanding of p-value and my overall confidence in my understanding of this essential concept - thanks, Zed!!🙏
That was exactly what I needed to understand this concept. I have watched so many videos trying to grasp this and only after watching your video do I understand. THANK YOU!!! You are a great educator.
This is the best video which explains p-value very clearly. I did watch many videos and lectures on p-value. But this one really make it clear !! Thanks Justin...!!
I am beyond grateful to you for taking the time to make such an amazing video for the public. I am in a five week accelerate statistic course this summer for grad school and it is rough doing it online, independently. I just came across your videos and I have tears of relief. Thank you!!!
I never understood what the p-value was, even after 3 courses where is used it. But somehow after the first 5 minutes of this video it was clear,. Thanks so much!
Hey, just wanted to let you know that these videos are extremely useful and the way each and every topic is explained is fascinating. Thank you very much for all the videos. Love from India.
Cool! I don't know much about stats, but the convention I think that is most conservative would be to choose a threshold; if doing a 2-tail test divide threshold by 2 and apply to both sides of the PDF, if doing 1-tail test, divide threshold by 2 still, and use half on one side... this is the most conservative way of doing it I guess... if I didn't divide by 2 in either case of a 2-tail or 1-tail test it would be less conservative with respect to the hypothesis.... but it means the 1tails are more conservative than the 2tails... I'm babbling.... Thanks for the video!
Hi Zed !! Great video. It's always a good thing to get a good intuitive understanding of the concept before diving into the mathematical details. It would be great if you could make a video on p-hacking or the misuse of p-values someday. Thanks!!
Appreciate you and this content so much! The structure (problem statement, context/history, charts, and application) make it easy to intuitively understand the concept. Thank you
Thanks for visualizing many these abstract statistic lessons but I’d like to see the picture of distribution curve in which the example that p-value is 1 between gender in both App and Control group.
Can you please explain How do you find out p values for different metrics in last example. I am assuming for age mean and other independent variables, you would possibly know the distribution from a large sample of data and use that distribution to calculate a p value. But what about the p value of those results how do u calculate that? How did u get distribution to calculate that
Can someone please explain where to work out the correct number to be divided by. For example 0.19 and .32 was used in this video? Thank you in advance.
@19:43 a clear stance on binarism, in a world of gender fluidity? Or pedagogical convenience? I may need to know how to take binary data and convert it to continuous to avoid discrimination...
Thank you so much Zed for your videos. It's a lot easier to understand when study with examples more than just a lot of theories. You are saving my life. Do you also teach about Time Series Analysis?
How did you calculate the p-value, I'm stuck. So, first we find the sd, which is root of 100(0.5)(1-0.5)= 5. Then, we convert to SND, (56-50)/5=1.2. And p-value= 2*P(Z>1.2)=2*.1151=0.2302 Could you please correct where I'm going wrong? Tysm, love you videos!
Provided our thereshehold 0f 0.05, hence is it ok to think that the mean diff should be in the shaded area (either one) for the two tailed and that the mean should be in te shades area in the one tailed in order for us to reject H0? Thanks! :)
Hi, I am from sri lanka .please can you help me to solve this problem , "You're planning a sample survey to estimate the proportion of registered patients fo particular hospital . You want to estimate p with 99% confidence and a margin of error no greater than 2% . Using a p of .5 ,how large of a sample do you need?"
Thanks for the video! I am a bit confused about why a confounder is sometimes a nonsignificant predictor based on likelihood ratio test. By definition a confounder has association with both - exposure and the outcome, and it's not in the causal pathway. So, if a regressor is a confounder, the p value of LR test on it's removal from the model should be significant, right? Why is it the case that regressors sometimes have nonsignificant p value of LR test on their removal from the model but are found to be a confounder as per change-in-estimate criteria? Is this because of type 2 error in LR test where we accepted the null hypothesis when we shouldn't have done so?
Why do you (Justin) call it an 'alternate' hypothesis? Alternating between what and what? Instead, should it not be called an 'alternative' hypothesis for it make sense exactly with what you are explaining?
OHHH MY GOD THE BEST VIDEO IN INTERNET THAT ACUALLY EXPLAINED P VALULE WITH REALLY SIMPLE EXAMPLES, I WOULD RECOMEND THIS TO ANYONE IF SOMEONE ASK ME! THIS VIDEO IS PURE GOLD! THANK YOU FOR POSTING THINGS THAT MAKE THE DIFFERENCE
Hi from Italy!
Finally I can write this very comment! I'VE PASSED MY STATISTICS EXAM :D
This is basically thanks to your very clear explanations! The way you break down concepts is just the best! Thank you very much for helping students all over the world.
This is a very late comment, but I case Zed still gets notifications: I’ve struggled with stats since college (40 years ago, ahem), but I’ve been trying to work on my understanding ever since. This vid has finally made p-values understandable. Thank you! Big fan.
Hi Justin! Just wanted to say I was in one of your tutorials for BUSS1020 in my first year of uni and you were honestly one of the best tutors I’ve ever had. You were always so passionate and have made me enjoy statistics so much more! Thanks so much and awesome video!
23:20 - "Eighteen percent probability that the difference we are seeing here occured purely by random chance"
Finally. I needed to hear these words. Understanding p-values was quite tough for me, I think i watched 10 video explainations but now it finally clicked in my head. Thank you very much
This is not correct, exactly. I repeated this many times and would like to involve @zedstatistics once and hear his take on it. I believe the correct interpretation would be, "The p-value of 0.18 means that if the null hypothesis (H0) is true, there is an 18% probability of obtaining results as extreme or more extreme than those observed on repeated testing. This is not the same as saying that there is an 18% probability of obtaining the observed effect by chance." Though I understand this is a minor thing since you're rejecting it anyway, for understanding, it's more like percentage vs percentile. It's more about the cumulative probability of extremes vs the probability of this exact finding.
Finally!! Someone who really knows what he's talking about AND can actually explain what p values are. Thank you sir, you rock!
This is amazing. I have ADHD and have always struggled with math in general. Statistics has been absolutely stressful for me during my MPH program. Thanks to you, I feel like I am finally understanding these concepts. THANK YOU!!!
Right there with you. It’s like I have to translate statistical terms in my head while my professor on 3 sentences ahead.
Finally understand p value better after watching most p-value youtube videos. This is the best!
Zed: I love your stats videos in general, but this one is through-the-roof informative - thanks so much!
I was trying hard over youtube for a good explanation of p values and finally i got the best ....
really helpful and clear explanation! I studied it 40 years ago and retrained in a graduate course again several years ago. I still feel unclear with how to interpretate P value. Great contribution!
The best content on p-value on RUclips, understood in one go!
Hi Zed! I'm Annet from Uganda, I have finally understood both P-value and Confidence Interval after watching your RUclips videos.
Thank you so much and Be Blessed.
I have to say this is the best and righteous video about P value, compare to a bunch of high-viewcount videos from blissfully misunderstood author about what p value is. Calculating the p value is not something complex, but the point that you really touch here is: Why rejecting the null hypothesis when the p value reach some small limit (5% to be common)? What is the link of trusting the null hypothesis with getting some really rare samples?
And i think you done a great job addressing this stuck point in understanding the p value. Great job. Keep it up!
I've watched numerous videos existing in youtube to understand p-values, and I can assert with no doubt that this one is the best explained. Thankyou
I have been looking for a simple explanation of the p-value for several days. Without a doubt, this is the best one of all! Thanks
Hi Zed ! Is it possible that one day you make a serie of videos where you just take a database and examine it juste like you’d do it as a statistician (for example, where do you begin, what questions do you ask yourself and how you answer them etc). I think it would be a perfect example that’ll show us how statisticians work ! Thank you for all your content
This. Is. Great. I'm definitely doing this. Will give you a big ol' shout out when I do it.
I support that request. It would enable us to apply all we have learned into practice! Thank you for your kind consideration.
Count me in too.
@@zedstatistics Thanks for all your content ! Eagerly waiting for this one!
@@zedstatistics did you do this?
This breakdown has thoroughly improved my understanding of p-value and my overall confidence in my understanding of this essential concept - thanks, Zed!!🙏
I watched 3 different videos on p values and still barely understood. This video helped SO MUCH and you totally crushed it dude tyvm
Honestly, I couldn't find a better video to explain this. THANK YOU SO MUCH!!!
I am finally truly understanding this concept now after many hours of frustration and despair! Thank you so much!
I have not seen any videos better than the yours. Lots of love from India. Kudos to your work.
That was exactly what I needed to understand this concept. I have watched so many videos trying to grasp this and only after watching your video do I understand. THANK YOU!!! You are a great educator.
Dude that's some sweet videos
i dont even need to know statistics to a deep level but your videos
still attract me, you are a great teacher!
6:15 - *I'm getting answer as 0.135626*, when I done sum(100Cx * .5^x * .5^(1-x)) where x = 56 to 100, why is that?
man, this semester I had a class in mathematical statistics. This was the one topic I didn't fully understand. THANK YOU!
AFTER SO LONG I ACTUALLY UNDERSTOOD p-value. Thankyou so much!!!!!
This channel deserves million subscribers.. Believe that!
This is the best video which explains p-value very clearly. I did watch many videos and lectures on p-value. But this one really make it clear !! Thanks Justin...!!
Thank the gods I found your channel a few months before my exam! Really enjoying your methods and delivery 😀 😍
Zed, you have no idea, but you've just managed to crack this nut for me. This is so well explained. Thank you, thank you....
You had me at "casting aspersions."
Excellent videos - thank you very much for your efforts.
I am beyond grateful to you for taking the time to make such an amazing video for the public. I am in a five week accelerate statistic course this summer for grad school and it is rough doing it online, independently. I just came across your videos and I have tears of relief. Thank you!!!
I never understood what the p-value was, even after 3 courses where is used it. But somehow after the first 5 minutes of this video it was clear,. Thanks so much!
These are the best stats videos... What a great way to teach, concept followed by examples. Wao
Hey Justin ..by far this is the best video and applicable for Med students on p value ..thanks a lot for all your effort
hey zedstatistics! thank you so much for explaning such a tricky topic in easy way. salute to you sir. you got a new subscriber.
Hi ! In 6:25 - How you have calculate as P-value as 0.193
incredible video. Wow 🤩 that was amazing and very helpful Justin
Thanks a lot from Saudi Arabia 🇸🇦
Excellent video. You make your video's easy to understand, interesting and you are a great speaker as well!
WE LOVE YOU MAN! waiting for your video series on full statistician work from a database. lead us master.
This is the best video about p-value! Deep respect! And thank you!
Thank you so very much! Hands-down the best lecture video on the topic that's up on RUclips!
My lecturer didn't explain at all. He just gave us the formula...
Very comprehensive and well explained. Thanks zedstatistics team.
recently discovered this..this channel is a gem, I'm in love
Really, thank you for the video! Not only did I learn about p-values, but it also made me look at statistics more fun rather than feeling stressful :)
your videos are lifesaving
Thank you, this is an excellent video, and as a fellow Aussie I enjoyed your use of the coin
Hey, just wanted to let you know that these videos are extremely useful and the way each and every topic is explained is fascinating. Thank you very much for all the videos. Love from India.
Thanks. Before this video, I could not wrap my mind around this concept of p-value.
Good examples and way to lead the viewer into understanding p- values.
Thank you!!! Thank you so much!!!! I was stuck in my Stats course - it just wasn't making sense to me until now.
Cool! I don't know much about stats, but the convention I think that is most conservative would be to choose a threshold; if doing a 2-tail test divide threshold by 2 and apply to both sides of the PDF, if doing 1-tail test, divide threshold by 2 still, and use half on one side... this is the most conservative way of doing it I guess... if I didn't divide by 2 in either case of a 2-tail or 1-tail test it would be less conservative with respect to the hypothesis.... but it means the 1tails are more conservative than the 2tails... I'm babbling.... Thanks for the video!
Hi Zed !! Great video. It's always a good thing to get a good intuitive understanding of the concept before diving into the mathematical details. It would be great if you could make a video on p-hacking or the misuse of p-values someday. Thanks!!
Hi Madhav! See the last 10 minutes of this video where I go into p-hacking! ruclips.net/video/kyjlxsLW1Is/видео.html
@@zedstatistics Cool. Thanks. Will check it out
thanks a lot..this gave me just enough understanding that I need to interpret the statistical data i often come across in publications
Awesome teaching. Really appreciate your efforts
Outstanding explanation of p-values! Thanks!
I am in Grad school, and these video really helping me catch up!
you made statistics cool, fun, interesting and understandable ! thank you sir, sincerely :)
Underrated channel
Love your French reading....Compliment from a French follower!
Best explanation I have seen. Thank you very much.
You are a great teacher... your videos are well worth the time. Thank you.
Cfr 6:10 What is on the Y-axis?
Appreciate you and this content so much! The structure (problem statement, context/history, charts, and application) make it easy to intuitively understand the concept. Thank you
You are a great teacher.
Thanks for visualizing many these abstract statistic lessons but I’d like to see the picture of distribution curve in which the example that p-value is 1 between gender in both App and Control group.
Wow...I'm recommending to my class, this helped me a ton!
Can you please explain How do you find out p values for different metrics in last example. I am assuming for age mean and other independent variables, you would possibly know the distribution from a large sample of data and use that distribution to calculate a p value. But what about the p value of those results how do u calculate that? How did u get distribution to calculate that
Your video's are so great!! can you do one about " time series analysis"??
Thank you, Mr. Justin.
Hey man, please, please make some videos on Multivariate Probability Distributions.
Can someone please explain where to work out the correct number to be divided by. For example 0.19 and .32 was used in this video? Thank you in advance.
You made something click finally ….thank you. I understand a better picture of what the P-value actually means. Beyond the Definition.
Thank you for your clear explanation
That's an excellent presentation....helped me a lot.... thank u very much ...
Hi, Thanks for all of this videos. they have been super useful to increase my understanding of hard statistics concepts
It is an interesting presentation but there is confusion b/n alph and P-value?
Can you please have a NAIVE BAYES Classifier session posted?
That was an exceptionally good introduction! Great content as always!
Ye Msc k liye seminar bna le kya
Ye konsi cllass ka h
@19:43 a clear stance on binarism, in a world of gender fluidity? Or pedagogical convenience? I may need to know how to take binary data and convert it to continuous to avoid discrimination...
Excellent video. Thanks!
This is sooooo helpful, Justin.
Thank you.
Thank you so much Zed for your videos. It's a lot easier to understand when study with examples more than just a lot of theories. You are saving my life. Do you also teach about Time Series Analysis?
Why aren't professors this good at teaching?
hello sir! can you make a video bout CONTINOUS PROBABILITY DISTRIBUTION AND JOINT PROBABILITY DISTRIBUTION
video tutorieal on non-linar least sqaure validaetion?
How did you calculate the p-value, I'm stuck.
So, first we find the sd, which is root of 100(0.5)(1-0.5)= 5.
Then, we convert to SND, (56-50)/5=1.2. And p-value= 2*P(Z>1.2)=2*.1151=0.2302
Could you please correct where I'm going wrong? Tysm, love you videos!
mee too ~ I am stuck here. Could someone give me some hints?
9:42 There should be twenty four "0", not twenty five. You move the decimal 25 times to the left.
excellent explanation
Woo, the explanation is so good, thanks so much
I still did not get the example of male and female baby born????
I am doing Chisquare test, Could you tell me how to find the p value range (a
Provided our thereshehold 0f 0.05, hence is it ok to think that the mean diff should be in the shaded area (either one) for the two tailed and that the mean should be in te shades area in the one tailed in order for us to reject H0? Thanks! :)
Please make some videos using Statistics related softwares. Like SPSS, E views.
Where we can apply these concepts practically.
Wow man, best explanation,thanx
Hi, I am from sri lanka .please can you help me to solve this problem ,
"You're planning a sample survey to estimate the proportion of registered patients fo particular hospital . You want to estimate p with 99% confidence and a margin of error no greater than 2% . Using a p of .5 ,how large of a sample do you need?"
Thanks for the video! I am a bit confused about why a confounder is sometimes a nonsignificant predictor based on likelihood ratio test. By definition a confounder has association with both - exposure and the outcome, and it's not in the causal pathway.
So, if a regressor is a confounder, the p value of LR test on it's removal from the model should be significant, right?
Why is it the case that regressors sometimes have nonsignificant p value of LR test on their removal from the model but are found to be a confounder as per change-in-estimate criteria? Is this because of type 2 error in LR test where we accepted the null hypothesis when we shouldn't have done so?
Why do you (Justin) call it an 'alternate' hypothesis? Alternating between what and what? Instead, should it not be called an 'alternative' hypothesis for it make sense exactly with what you are explaining?