Power Calculations in Hypothesis Testing | Statistics Tutorial #17 | MarinStatsLectures

[@marinstatlectures]

👋🏼 Hello there! In this video we explore how the statistical power is related to alpha, the sample size (n), and the difference we wish to detect. We also go over the calculations for statistical power and errors in hypothesis testing. If you like to support us you can Donate (bit.ly/2CWxnP2), Share our Videos, Leave us a Comment and Give us a Like 👍🏼 ! Either way We Thank You! 🦄

[@satishhulkund6841]

How can equality be in both null and alternative hypothesis

[@somdubey5436]

This video is a gem when it comes to understanding Power of a Test. Have not came across any such other informative video till now. Thanks a lot.

[@Plankton_Doctor]

This is great. I have been hearing about power for a very long time and read a little about it, but this makes it much more clear. Thanks!

[@marinstatlectures]

great to hear!

[@jiesun31]

this is the clearest video on the power

[@Emi_Cooks12]

This is much easier to understand!!! I have been struggling with this for weeks. Thanks

[@viethuynh5453]

One of the best stat lecture. I hate so much with the formula all stat course show us.20 years still confuse and forgot even I had revisit stat over 50 times

[@jimmyodiazportillo6978]

I understand a bit of English only, but this is an spectacular video, I would like all profesors in colleges were like you. Thank you so much for this videos so important.

[@sudhakarmadakaful]

while calculating Beta at 12:12, for calculating, z- static given the Ha is true, why did you use Std. Dev. of population, wouldn't it be different for new distribution if Ha is true?

[@waqarulmulk3530]

very well explained 👏

[@firstkaransingh]

Thank You. This was very helpful

[@marinstatlectures]

you're welcome!

[@zoelin1891]

THANK YOU!!!!!!!!!!!

[@agneskyamulabi9474]

Hi Mike, where do you get the 1.65 from in this video?

[@happygrace8065]

Thank you!! 💫

[@mehrdadanvar8518]

beautifully explained.

[@a.c.6416]

Thanks a lot, your explanations are just on point! It helped me a great deal :)

[@markkennedy9767]

Why is the alternative hypothesis (mu

[@simonepost5937]

Hi, I might be wrong but I though that at 5:49 the s in the Zstat should be a sigma because the s is for the t test and sampling distribution but now the Standard deviation is known it should bee a sigma right? You know what I mean? For the rest really helpful but I was just confused

[@buihung3704]

Yep, it was supposed to be the sigma, what he tried to indicate that using t-score is more appropriate than z-score but he used z-score for the sake of simplicity. But when sample size gets larger, it doesn't matter anymore :3

[@poonambajaj17]

Very well explained but little bit tough. Our teachers did not explained in such a good sense. I am student of statistics department. It took a little bit time even for me to understand. I appreciate the work you done. Kindly explain how to choose sample size for different study designs?

[@joselev271]

Hey Marin, thank you very much for your videos, they are awesome, great, neat, andvery helpful. I have just one QUESTION: Where do you Explain the t table type of exercise? I hear you say we do it with z to simplify, but i wanted to know about t. THank you very much!!!

[@Cuicui229]

Thank you:). it helps me a lot!

[@RobvanMechelen]

Dear Martin, thank you for this lecture. I still have a question : Power = P(reject H0 | Ha =True) as you state it. Given that Ha = true, then the mean BP = 145 mm Hg +/- SD for people using the drug A power of 80% tells us that the chance = 80% that we observe a mean blood pressure = 145 mmHg +/- SD in these people, but there still is a 20% chance that their mean blood pressure still = 160 mmHg +/- 36? Is this the correct translation of power is P(rejecting the H0|Ha is true)?

[@joaoguilhermegiancursitedd8988]

Why did you use the std deviation of the original population for calculating beta zstat? Wouldn’t the alternative distribution have a different std deviation? Thanks for your video, the best i’ve ever watched

[@marinstatlectures]

That would be our best guess at the SD. And most often we’d assume the SD to be a certain value regardless of the value of the mean. Hope that made sense

[@cyanide4u539]

Can you add some lectures about Nonpaametric tests?????? I need help in that

[@marinstatlectures]

Coming up... we’re going to be ready to post one about the Wilcoxon Signed Rank test within 1 week. We also use this topic to introduce the general concept underlying a non parametric test

[@cyanide4u539]

Thank you very much for consideration, please upload Kulmogrov test and Brown -Moods method for regression too, looking forward for videos on non parametric statistics too. Thanks Again

[@poonambajaj17]

Please make videos on moderator data and cluster analysis based on real dataset

[@marinstatlectures]

Thanks for the suggestion. Right now I mostly create videos related to the courses I’m teaching

[@namduki]

Thank you, if i knew you, i would not have had a D in statistic in University :(

[@hime6000]

Why did you do "

[@marinstatlectures]

That is a one-sided alternative

[@EL-eo8dh]

👍

[@dayakarreddy9266]

May i know how u got z score 1.65 ?

[@Cuicui229]

Hi I am learning the videos too. He mentioned that it is based on the z distribution table alpha=0.05.

[@dayakarreddy9266]

Lu Zhang thanks !!!