The thing I hate about professors is that they are somehow able to always explain the most simplest of concepts in the most confusing of terms. Thank you for this clear explanation.
Honestly, I think they do this because some of them want to feel important. Einstein himself said, "If you can't explain (the subject) simply... then you don't know it well enough yourself..."
Having my statistics exam next week and I can say you just saved my life. I have been struggling to get this since the beginning of the semester. You just put it up to me in the most astute way everr
Miss, I'm literally about to take exam on this material...😭😭 You explain this much better than my professor. Thank you very much for this, I now understand that the confidence level correlate directly with the Z-level which I can use to calculate the confidence interval. From the bottom of my heart, I sincerely thank you😭😭
This video provides an all too common misinterpretation of what a confidence interval is. 95% confidence interval refers to the procedure in which the CI was created, NOT the probability that a given parameter values lies within a given confidence interval. 2:37: 'If a confidence interval of 95% is given, one can be 95% sure that the true parameter value lies within this interval' That is simply wrong. Frequentist statistics (probability estimates derived from repeated (long-run) measurements from sampling) cannot provide probability estimates on model parameter values, this is a feature of Bayesian statistics. Confidence intervals do not provide a statement about the parameter as it relates to any given sample at hand. It provides a statement of an average performance of the procedure (long-run performance) of drawing intervals that 95% are expected to contain the true parameter value. You can not however, know if any given (random) sample CI will contain the true parameter value. A given CI either does, or does not contain the true parameter value.
Yes, thank you for writing this. For anyone who would like to learn more about this, consider reading “The fallacy of placing confidence in confidence intervals” (DOI: 10.3758/s13423-015-0947-8) or E. T. Jaynes’ “Confidence Intervals vs. Bayesian Intervals” (1976).
Your accent brings my memory back to my Econometrics professor who is from Eastern Europe. She taught lots of statistics that's very hard to understand :)
Thank you so much for your explanations, which are excellent!I Most appreciate you sharing this content for free for all of us trying to get our head around statistics! Cheers!
Thanks for the video, but I think i've spotted a small mistake, since you are using the sample sd you shold base your confidence interval on the t-distribution, not on the nomal distribution, exept, when your sample size is large, then z and t values are similar.
Hi many thanks for your question! No, that is exactly the purpose of the confidence interval. For example, if you take the 95% confidence interval, it is 95% likely that the mean of the population lies within the confidence interval and 5% likely that it does not lie within the confidence interval. Since we do not know the true value, we can only estimate and give a probability.
What the video describes is _not_ a confidence interval, but instead a (Bayesian) credible interval. Only the latter allows the interpretation (at 2:33) that "one can be 95% sure that the true parameter lies with this interval." A single confidence interval by itself allows no such interpretation, only a large collection of them from many repetitions of the same experiment.
Well done! Usually I do not like to comment but I feel like I have to do it now. I have issues with Statistics and soon I have an exam. This video helped me a lot. Thanks!
Could you explain the s/sqrt(n) part. I guess that std. Deviation equals the sqrt(variance divided by n). So why is there yet another division with n when you calculate the C.I.?
If you divide by sqrt(n), you calculate the standard deviation of the mean (Also called standard error), that is, how much the mean varies. Without sqrt(n) you calculate how much the whole data varies. But we want to know in which range the mean fluctuates so wie use s/sqrt(n). Regards Hannah
i too, but i researched and the answer is we need to calculate area to the left first, we look for that number in the z table . here is the video that helped me www.google.com/search?q=how+to+read+z+table+confidence+interval&rlz=1C1CHBF_enIN1043IN1043&oq=how+to+read+z+table+confidence+inte&gs_lcrp=EgZjaHJvbWUqBwgCEAAYgAQyBggAEEUYOTIGCAEQRRg7MgcIAhAAGIAEMggIAxAAGBYYHjIKCAQQABiGAxiKBTIKCAUQABiGAxiKBdIBCTEyNDgwajBqN6gCALACAA&sourceid=chrome&ie=UTF-8#kpvalbx=_0bH9ZN_eD4rsseMPmJGwsAE_28
It is more usual to call the sample mean "x bar". It is also more usual, when calculating confidence intervals of means from a normally distributed population, to use the student-t critical value rather than the Z, particularly when you are using the sample std deviation. The formula you are showing here would either be an approximate confidence interval or a large sample confidence interval for the population mean. Also, be very careful when interpreting confidence intervals. In classical (frequentist) statistics, it is not true to say "the population mean will fall within the confidence interval 95% of the time". In classical statistics we assume population parameters are fixed, unknown constants, so apportioning probability to their value is incorrect. The confidence interval is the random quantity here. If you want to correctly interpret a 95% frequentist CI, you need to acknowledge it is the CI that varies randomly from sample to sample and say something like "if 100 samples were taken from this population, and a 95% confidence interval (for the mean) was calculated each time, 95 of these intervals would contain the true population parameter (mean)." If you want to interpret interval estimates in the way you have described in your video, you need to become Bayesian and talk about credible intervals.
Nice video , I have a question. assuming I have a non normal population and only one sample with its respective statistics, can I assess this confidence interval there? I mean , according to your example I need more than one sample to get a mean of the sample distribution . Please that would really help me
how can you assign a probability to the population mean that is constant in frequentist statistics? μ is either within or outside the 95% CI it does not have an associated probability. Right?
That was a brilliantly crafted explanation for the concept behind a Confidence Interval. Thank You ! If someone can help me explain how the z-value for a certain High Probability (for instance 1.96 for 95% probability) is ascertained in depth, then that would be wonderful. Thanks,
Calculate (integrate) the area under the curve of the standard normal distribution (mu=0, sigma=1) in such a way, that area is 0.95 from -inf to x. Since there is no explicit form of that integral people botherd to calculate it for "all" (aka 10'000) values between 0 and 1 so it can be printed to a book and looked up.
Thank you. I have a Q. When you take a sample and calculate itsmean, do u put that sample back and let it's elements mix with population and take another sample or do u keep first sample aside and take another sample?
If you like, please find our e-Book here: datatab.net/statistics-book 😎
The thing I hate about professors is that they are somehow able to always explain the most simplest of concepts in the most confusing of terms. Thank you for this clear explanation.
Honestly, I think they do this because some of them want to feel important. Einstein himself said, "If you can't explain (the subject) simply... then you don't know it well enough yourself..."
😂😂😂
Having my statistics exam next week and I can say you just saved my life. I have been struggling to get this since the beginning of the semester. You just put it up to me in the most astute way everr
Miss, I'm literally about to take exam on this material...😭😭 You explain this much better than my professor. Thank you very much for this, I now understand that the confidence level correlate directly with the Z-level which I can use to calculate the confidence interval. From the bottom of my heart, I sincerely thank you😭😭
Many thanks again for your nice feedback!!! Good luck for you exam! Regards Hannah
So...How'd it go?
How did you do on the test?
@@queenof2kings299 She failed, I'm in her class
@@danielr7008 😳
Thank you for explaining this better than my university
Thanks! : )
True
in 5 minutes 😁
Retweet
Thanks
Lady you are single handedly saving my medical degree ily
Greate : )
Watching this 4 hours before finals! Pure magic ✨
Thanks!
Who has made the video lecture is a real professional of their work. Not only easy to understand but interesting to watch in some ways!!!
Many many thanks!!!!
You are so talented. Your explanations are clear, concise, direct and accessible. Lovely.
Whenever I start to study for an exam I always look at youtube videos for a spesific topic. Very well explained, thank you for video.
Glad it was helpful!
This earth needs more teachers like you, thnak you for explaining!
Many thanks!
I have seen plethora of videos on statistics. But your views are explicit and conspicuous. Your method of teaching is comprehensive.
Wow, thanks : )
Re-teaching myself statistics after 15 years is not easy... THANK YOU!
Greate! Regards Hannah
I can't express how happy I am that I found your videos. You make statistics understandable!! Thank you! Thank you! Thank you!!
Happy to help and many thanks for your nice feedback! Regards Hannah
Your explanation is way better and easier than I expected TMY of this site.😍
Glad to hear that
This video provides an all too common misinterpretation of what a confidence interval is. 95% confidence interval refers to the procedure in which the CI was created, NOT the probability that a given parameter values lies within a given confidence interval.
2:37: 'If a confidence interval of 95% is given, one can be 95% sure that the true parameter value lies within this interval' That is simply wrong.
Frequentist statistics (probability estimates derived from repeated (long-run) measurements from sampling) cannot provide probability estimates on model parameter values, this is a feature of Bayesian statistics.
Confidence intervals do not provide a statement about the parameter as it relates to any given sample at hand. It provides a statement of an average performance of the procedure (long-run performance) of drawing intervals that 95% are expected to contain the true parameter value. You can not however, know if any given (random) sample CI will contain the true parameter value. A given CI either does, or does not contain the true parameter value.
Yes, thank you for writing this. For anyone who would like to learn more about this, consider reading “The fallacy of placing confidence in confidence intervals” (DOI: 10.3758/s13423-015-0947-8) or E. T. Jaynes’ “Confidence Intervals vs. Bayesian Intervals” (1976).
Completely agree. Viewers, beware.
Beautifully explained! Short, crisp and easy to understand. Thanks a lot!
Glad it was helpful!
Your accent brings my memory back to my Econometrics professor who is from Eastern Europe. She taught lots of statistics that's very hard to understand :)
: )
You are the only channel that clearly explains this topic, you make it so easy to understand. Thank you :)
Glad you liked it and many thanks for the nice feedback! Regards Hannah
Thanks teacher
This was one of the most useful videos on confidence intervals. Thank you so much.
I don't know what to say but you make it so simple i really appreciate your efforts
Thank you so much 😀
Thankyou so much, never stop uploading please.
Never seen such a crystal clear examination. Thank you so much❤
Thank you so much for your explanations, which are excellent!I Most appreciate you sharing this content for free for all of us trying to get our head around statistics! Cheers!
Glad it was helpful and many thanks for the nice feedback! Regards Hannah : )
They explained this so easy it’s annoying. Great content!
Thank you for simplifying the concept to a level we can understand
Glad it was helpful!
Explained so well, it has cleared my concept without looking into the book
Madam, thanks a ton. Your presentation is simply superb.
This is the best explanation of confidence interval on all of youtube (at least the first 10 search results)
Many many thanks for the nice feedback!!! Regards, Hannah
I agree. The rest are filled with Phd Professors complicating it more lol
@@chvrleshd Many thanks : ) Regards, Hannah
I could never understand this before but this really helped me!!!❤
I truly love your channel. Keep doing the best work.
Thanks for the video, but I think i've spotted a small mistake, since you are using the sample sd you shold base your confidence interval on the t-distribution, not on the nomal distribution, exept, when your sample size is large, then z and t values are similar.
Hi, many many thanks for the hint! that is of course true! Regards Hannah
you're far better than my lecturer
Many thanks!!!
Thankksssssss... no word to thank you... just can say God bless you...
Most welcome 😊
Really fascinating explanation keep your journey up 😊
Excellent, wonderful and easy to use explanation. Congrats!!
Really great video , imma use this for my exam . Thanks a lot
Glad it was helpful!
Really I appreciate the way you explain it in simple and understandable way. Thank you so much
So nice of you
Very well explanied and nice presentation, Thank you !
Excellent explanation. Thank you!
Thank you, this is so well explained - life saver!!
Glad it helped!
Incredibly easy to understand 👍
Glad to hear that! Regards Hannah
If you don’t know the value of the population mean, is it possible to know *for* *certain* if the confidence interval contains the population mean?
Hi many thanks for your question! No, that is exactly the purpose of the confidence interval. For example, if you take the 95% confidence interval, it is 95% likely that the mean of the population lies within the confidence interval and 5% likely that it does not lie within the confidence interval. Since we do not know the true value, we can only estimate and give a probability.
Really underrated video.
I appreciate that!
Thank u my friend From GREAT OROMIA ETHIOPIA 🇪🇹 💖 💓
What the video describes is _not_ a confidence interval, but instead a (Bayesian) credible interval. Only the latter allows the interpretation (at 2:33) that "one can be 95% sure that the true parameter lies with this interval." A single confidence interval by itself allows no such interpretation, only a large collection of them from many repetitions of the same experiment.
Thanks so much this is so meaningful to me❤
Very well explained! Thank you!
Glad it was helpful!
You're jsut fantastic in your explanations!
Glad it was helpful! Thanks for your nice feedback! Regards Hannah
it's so perfect
how we use pre/post test in DATAtab
Many thanks for your feedback! I will put it on my to do list!
appreciate your timely response
May I have your email address.
Well explained mam. Thank you!
very useful video. you helped me alot!!
Glad to hear that!
A world of thanks for this video.
So nice of you
Excellent , I like the discription easy to visualize and understand.
This was very, very helpful!
This is very helpful -sir adib student
Well done! Usually I do not like to comment but I feel like I have to do it now. I have issues with Statistics and soon I have an exam. This video helped me a lot. Thanks!
Many many thanks!!!!
Thank you so much for the explanations
Best explaination I've seen
day after tomorrow is my xm & u saved meeeeeeeeeeeeeeeeeee
Thanks : )
Could you explain the s/sqrt(n) part. I guess that std. Deviation equals the sqrt(variance divided by n). So why is there yet another division with n when you calculate the C.I.?
If you divide by sqrt(n), you calculate the standard deviation of the mean (Also called standard error), that is, how much the mean varies. Without sqrt(n) you calculate how much the whole data varies. But we want to know in which range the mean fluctuates so wie use s/sqrt(n). Regards Hannah
It is a useful class..thank u sweet heart.
You are welcome 😊
For that standard of error, I learned that it’s used with the T-score and not the Z. The SE you would use is sqrt(π (1-π)/ N or population sample)
hi, can you please tell me how did you determine na z as 1.96 from the table? I am getting z as 1.65 :(
i too, but i researched and the answer is we need to calculate area to the left first, we look for that number in the z table . here is the video that helped me www.google.com/search?q=how+to+read+z+table+confidence+interval&rlz=1C1CHBF_enIN1043IN1043&oq=how+to+read+z+table+confidence+inte&gs_lcrp=EgZjaHJvbWUqBwgCEAAYgAQyBggAEEUYOTIGCAEQRRg7MgcIAhAAGIAEMggIAxAAGBYYHjIKCAQQABiGAxiKBTIKCAUQABiGAxiKBdIBCTEyNDgwajBqN6gCALACAA&sourceid=chrome&ie=UTF-8#kpvalbx=_0bH9ZN_eD4rsseMPmJGwsAE_28
Thanks so much for your clear explanation
Thanks!
It is more usual to call the sample mean "x bar". It is also more usual, when calculating confidence intervals of means from a normally distributed population, to use the student-t critical value rather than the Z, particularly when you are using the sample std deviation. The formula you are showing here would either be an approximate confidence interval or a large sample confidence interval for the population mean.
Also, be very careful when interpreting confidence intervals. In classical (frequentist) statistics, it is not true to say "the population mean will fall within the confidence interval 95% of the time". In classical statistics we assume population parameters are fixed, unknown constants, so apportioning probability to their value is incorrect. The confidence interval is the random quantity here. If you want to correctly interpret a 95% frequentist CI, you need to acknowledge it is the CI that varies randomly from sample to sample and say something like "if 100 samples were taken from this population, and a 95% confidence interval (for the mean) was calculated each time, 95 of these intervals would contain the true population parameter (mean)."
If you want to interpret interval estimates in the way you have described in your video, you need to become Bayesian and talk about credible intervals.
i love how all the good explanation videos only come after i did my exam 😭
Many thanks : )
You are a life saver, Many thanks upon thee.
Many thanks!!! 😊 Regards, Hannah
Great job! Simply explanation
Thank you. you were easy to follow and I'm feeling empowered.
Can you elaborate on the number ranges for a CI? What does it mean if it includes 0, doenst include 0, less than 1, higher than 1?
It was absolutely fantastic explanation.
But, a question...why mostly it is calculated at 95%. Why not 98 or 99% 🤔
Excellent explanation. Thank you.
Glad it was helpful and many thanks for the nice feedback! Regards Hannah
Thank you… Very well done 👍🙏
Nice video , I have a question. assuming I have a non normal population and only one sample with its respective statistics, can I assess this confidence interval there? I mean , according to your example I need more than one sample to get a mean of the sample distribution . Please that would really help me
You're a life savior
Very helpful thank u mam
Most welcome 😊
Your explanation is so clear thank you!
Glad it was helpful!
This was an excellent video !! Thankyou !
Glad you enjoyed it and thanks for the Feedback!
Lovely.. Well explained
Glad you liked it
You're a life saver. Thank you so much for your videos 🙌
Great explanation ! , very easy to understand appreciate the time you put into this !
Very well explained, thanks!
4:45 HOW ON EARTH did you get 1.96??? My table says that's a value of .9750. Shouldn't the z be 1.65 where it's value is .9505??
Check that you'rd not looking at a 1 tailed test (therefore 95% is to the left) vs. a 2 tailed test (where 95% is around the mean)
@msralmjs i just learned this in my stat class yesterday ;-;;;; thanks for tellin me man
Thank you queen!!! Helps a lot!
You're so welcome!
very well explained and easy to grab the idea
Thank you 🙏. Now I get it so well..😊😊
Great job. Well explained. Thanks
Many thanks! 😊Regards, Hannah
......you are good. Very GOOD, Excellent.
Thank you very much!
Well presented. Thank you so much!
Wow that was really well explained
Many thanks for your feedback! Regards, Hannah
Very didactic teaching! Thank you very much!
how can you assign a probability to the population mean that is constant in frequentist statistics? μ is either within or outside the 95% CI it does not have an associated probability. Right?
Well explained. Thank you 👏🏽
OMG I am.lucky I found you ❤️❤️
Many thanks for your nice feedback!!! Regards Hannah
That was a brilliantly crafted explanation for the concept behind a Confidence Interval. Thank You !
If someone can help me explain how the z-value for a certain High Probability (for instance 1.96 for 95% probability) is ascertained in depth, then that would be wonderful.
Thanks,
Calculate (integrate) the area under the curve of the standard normal distribution (mu=0, sigma=1) in such a way, that area is 0.95 from -inf to x. Since there is no explicit form of that integral people botherd to calculate it for "all" (aka 10'000) values between 0 and 1 so it can be printed to a book and looked up.
Very well explained !!
thank you for sharing 💕💕💕💕
Many thanks! Very clear and well explained!!!
Glad it was helpful!
Thank you. I have a Q. When you take a sample and calculate itsmean, do u put that sample back and let it's elements mix with population and take another sample or do u keep first sample aside and take another sample?
Thank youuuuu sooo muchhhh❤❤❤❤❤