Statistics 101: Calculating Type II Error, Concept with Example

Statistics 101: Calculating Type II Error - Part 1
Part 1: Conceptual Background with Example
Part 2: Curve Animation and Test Power
In Part 1 of this video, we learn how to find the level of Type II error in a single sample hypothesis test. Type I error is easy...we chose it. However, Type II error is elusive. We examine the distribution differences graphically and then use an example problem to determine various levels of Type II error. Finally, we discuss hypothesis interpretations given certain sample mean locations. Enjoy!
My playlist table of contents, Video Companion Guide PDF documents, and file downloads can be found on my website: www.bcfoltz.com

Пікірлер: 94

  • @Lexing187
    @Lexing1878 жыл бұрын

    This is how every teacher/professor needs to teach. Introduce something with a conceptual back ground (the purpose behind the topic) and then go into examples. Too many times they try to teach by example. Thank you for these wonderful videos!

  • @BrandonFoltz

    @BrandonFoltz

    8 жыл бұрын

    +Lexing187 You are very welcome! Thank you so much for watching and please share. Hang in there!

  • @harshalshinde227

    @harshalshinde227

    8 жыл бұрын

    Those who are watching more videos of urs, ur inspirations sounds boring after some time, plz have a tab for skipping intro, thanx inadvance.

  • @agermoune

    @agermoune

    7 жыл бұрын

    I know you are entitled to your own opinion, but if you can't say something constructive please keep your mouth shut. encouraging words never get boring, besides do some research on how to forward on youtube. Ungrateful!

  • @jmhome2267
    @jmhome22676 жыл бұрын

    I wish I had you as my teacher, tutor, and professor throughout my education of statistics. Thank you for doing this!

  • @julianasilvadedeus574
    @julianasilvadedeus5744 жыл бұрын

    This was one of the hardestest topics I've ever studied in stats. But thanks to your advices, a kept my head up and passed through this temporary rough patch! Thank you for teaching stats and perseverance.

  • @gentle2005phir
    @gentle2005phir7 жыл бұрын

    I bow to your explanatory skill. You are doing great service.

  • @BrandonFoltz

    @BrandonFoltz

    7 жыл бұрын

    Thanks so much!

  • @Alexjames969
    @Alexjames9694 жыл бұрын

    I am a grad student and was struggling to understand this concept. I felt embarrassed not knowing this as an engineering student. Thank you so much

  • @adhiththyagarajan8712
    @adhiththyagarajan87126 жыл бұрын

    I wish you could teach every subject that I have been taking up! Would make my life a lot easier and happier. Thanks a lot for your videos. :)

  • @fu-changsun923
    @fu-changsun9236 жыл бұрын

    One of the BEST teacher/lecturer who can effectively convey the knowledge! Thanks a lot.

  • @mmsas1
    @mmsas19 жыл бұрын

    The best explanation I've come across, so far. Thank you.

  • @MeoWooff_01
    @MeoWooff_013 жыл бұрын

    Very clear explanation. I have seen several videos on this topic. But this has the clearest step-by-step explanation.

  • @vipingarg1349
    @vipingarg13496 жыл бұрын

    Best & most clear explanation on this concept I have come across! Thank you for the exceptional video.

  • @maneeshnair79
    @maneeshnair795 жыл бұрын

    Excellent. I was struggling to understand the Type 2 error calculation. Very well explained, especially with the stress on key points twice. Thanks!

  • @sruthichiramel
    @sruthichiramel5 жыл бұрын

    You are the most amazing Stat Teacher . God Bless you!

  • @Patricia-zi3il
    @Patricia-zi3il4 жыл бұрын

    THANK YOU FOR DEMOCRATIZING KNOWLEDGE IN SUCH AN AMAZING WAY

  • @sanjaykrish8719
    @sanjaykrish87194 жыл бұрын

    I was soo confused of this concept.. but after seeing this video I got it very very clear. Thanks a ton

  • @zhenfeiding5945
    @zhenfeiding59455 жыл бұрын

    his teaching is awesome! definitely better than my grad school teacher in the Columbia University

  • @informationhub5548
    @informationhub55486 жыл бұрын

    Extra ordinary art of communication

  • @RobvanMechelen
    @RobvanMechelen8 жыл бұрын

    Again, Brandon a superb presentation on the type II error. You made it also perfectly clear in previous presentations that a type I error (alpha) often is a nuisance (fire alarm example and toyota car example) whereas a type II error (beta) can be catastrophic. Why is it in power calculations that one is permissive for type II (beta) errors (10-20 %) and rigorous for type I (alpha) errors (1-5%). Why not the opposite ?

  • @francesmcbrien4785
    @francesmcbrien47852 жыл бұрын

    Thank you for these videos. As a much older person back in school, when instructors assume the math and don't show it, I get lost. I'd like to see every part of each equation, please. In addition, when you go to a chart and get a number (which is probably assumed by all the younger students) please let me know that you have done that because I get lost there as well. Thanks so much. Your encouraging words prior to the video helped me. xo

  • @R99999S
    @R99999S8 жыл бұрын

    Great work! Keep it going!

  • @rafaelrodsan1
    @rafaelrodsan14 жыл бұрын

    Amazing video, I never found a course so well explained, Thank you Bradon

  • @pratikedu8331

    @pratikedu8331

    3 жыл бұрын

    when the population mean is assumed to be 4 you mentioned there is a 0.16 probbablity of not rejecting the null hypothesis when it should have been rejected, my question is why should we have rejected the null hypothesis with a population mean of 4 when it falls in the brown region. anyone if understood my query please help me understand.

  • @aries284769
    @aries2847697 жыл бұрын

    Did you have any videos for Calculus? I am keen to see them coming from you..Thanks

  • @huilu8637
    @huilu863710 жыл бұрын

    your video is amazing and helpful. thank you very much

  • @mbuvigodfrey2335
    @mbuvigodfrey23359 жыл бұрын

    very well explained, thank u

  • @edwardgooden9875
    @edwardgooden987510 жыл бұрын

    Much clearer. Thanks.

  • @nathannzau7490
    @nathannzau74909 жыл бұрын

    Great explanation mate. Keep it up

  • @TheHynza
    @TheHynza11 жыл бұрын

    thanks a lot. i have a midterm 4 hours later, it helped a lot.

  • @jonathanx4540
    @jonathanx4540 Жыл бұрын

    Wonderful explanation thank you.

  • @henriquesicoli4644
    @henriquesicoli46447 жыл бұрын

    you are awesome Brandon. All the best men

  • @EngineeringChampion
    @EngineeringChampion3 жыл бұрын

    You make statistics so easy! Thank you!

  • @pratikedu8331

    @pratikedu8331

    3 жыл бұрын

    when the population mean is assumed to be 4 you mentioned there is a 0.16 probbablity of not rejecting the null hypothesis when it should have been rejected, my question is why should we have rejected the null hypothesis with a population mean of 4 when it falls in the brown region. anyone if understood my query please help me understand.

  • @reubenchemmanam2502
    @reubenchemmanam25024 жыл бұрын

    great vid! thanks brandon!

  • @ThankYouAYODHYA
    @ThankYouAYODHYA5 жыл бұрын

    PERFECT.... !!! Thank You.

  • @TanyaDesai
    @TanyaDesai11 жыл бұрын

    I have no words. Thank you. I might not fail after all :D

  • @thanhvannguyen5787
    @thanhvannguyen57876 жыл бұрын

    Hi brandon, thanks for your efforts, I was wondering if you could share your powerpoint so that i could use it for reviewing lessons

  • @sajasadek5541
    @sajasadek55413 жыл бұрын

    Thank you for explaining the concepts really good. I just got a question about if there was a way to find 0.16 and 0.84 manually without using Excel. I am revising for a test and the lecturer said we are only using pen and paper to calculate Type II error.

  • @brandochouthi
    @brandochouthi5 жыл бұрын

    Great video!

  • @kiranphaniharam3859
    @kiranphaniharam38598 жыл бұрын

    great explaination sir

  • @rajarshi403
    @rajarshi4037 жыл бұрын

    You sir are a rockstar!

  • @migueltoralla2442
    @migueltoralla24423 жыл бұрын

    Terrific explanation! Hope you get tons of views cuz you deserve them!

  • @BrandonFoltz

    @BrandonFoltz

    3 жыл бұрын

    Much appreciated!

  • @thryce82
    @thryce823 жыл бұрын

    excellent video this confused me for weeks

  • @pratikedu8331

    @pratikedu8331

    3 жыл бұрын

    when the population mean is assumed to be 4 you mentioned there is a 0.16 probbablity of not rejecting the null hypothesis when it should have been rejected, my question is why should we have rejected the null hypothesis with a population mean of 4 when it falls in the brown region. anyone if understood my query please help me understand.

  • @dnlsn7879
    @dnlsn78795 жыл бұрын

    As with all your videos, great explanation! I am learning to love statistics which I hated at uni. Unfortunately at that time these type of videos didn't exist. Hope somehow you get rewarded for this great work. However for this particular video, shouldn't you have used s=1.5 as it is based on the sample?

  • @sunnypathak2335
    @sunnypathak23358 жыл бұрын

    You have virtual way...........wow I have understood much

  • @kritiohri2950
    @kritiohri29504 жыл бұрын

    Thank you for the video, sir do you have the video on how to find type I error probability?

  • @joejohnoptimus
    @joejohnoptimus7 жыл бұрын

    To find the beta probability (type II error) when the hypothesis test is testing for the difference in two sample proportions (not testing for a single mean), would we just use this same process?

  • @jaqoffkuntz
    @jaqoffkuntz7 жыл бұрын

    The beginning of the video made me want to cry. Thank you... so positive

  • @wfalcao69
    @wfalcao695 жыл бұрын

    AWESOME !!!!!!!

  • @Sunshine-eo7mb
    @Sunshine-eo7mb4 жыл бұрын

    thank you!

  • @TimvanWessel
    @TimvanWessel8 жыл бұрын

    Hi Brandon! I have watched playlist 9 completely and one thing is unclear to me from fully understanding this PL. When do i know that i am making a type I or II error? At 19:30 in this video there is A, B, C & D. but how do I know where these values belong at µ0 or µa? (If they weren't in the graph but in a list like "A=2.5, B=3.8, C=3.4 & D=4.2") . Maybe it is a stupid question and I have missed something but I don’t grasp this part :(.

  • @karanyadav373
    @karanyadav3739 ай бұрын

    thank u so much :D

  • @tonisams6902
    @tonisams69027 жыл бұрын

    Thank you.

  • @SY-fb7yc
    @SY-fb7yc7 жыл бұрын

    Since sample size is 25, shouldn't we use the t- test instead of z- test?

  • @TheDeividasLTU

    @TheDeividasLTU

    5 жыл бұрын

    He has the standart deviation

  • @mrei7119
    @mrei71194 жыл бұрын

    Thank you so much for the video. However, if you don't mind, I would like to ask Mr. Foltz (or anyone who's kind enough) the following question: At approximately 9.25 min, you explain that to find the type II error, the population mean of the alternative hypothesis has to be greater than 3 (e.g. 3.5, 3.7, 4 etc.). This is the part where I find weird. Why do we have the freedom to choose the population mean of the alternative hypothesis in the first place? If we have that freedom, then the values of possible type II errors seem to be infinitely many. Then, how is finding the type II error be practical or useful?

  • @mollypete4270
    @mollypete42706 жыл бұрын

    Vielen Dank!!!!!!!!!!!!!!!!!!!!!!!!

  • @prithvishbanerjee6922
    @prithvishbanerjee69223 жыл бұрын

    you are a freakin genius man!

  • @pratikedu8331

    @pratikedu8331

    3 жыл бұрын

    when the population mean is assumed to be 4 you mentioned there is a 0.16 probbablity of not rejecting the null hypothesis when it should have been rejected, my question is why should we have rejected the null hypothesis with a population mean of 4 when it falls in the brown region. anyone if understood my query please help me understand.

  • @octaviodelsol4998
    @octaviodelsol499810 жыл бұрын

    thank you!!!!! i was stuck there for a while

  • @BrandonFoltz

    @BrandonFoltz

    10 жыл бұрын

    You are very welcome! :) Take care, B.

  • @octaviodelsol4998

    @octaviodelsol4998

    10 жыл бұрын

    (Y) good job!

  • @prabhudaskamath1353
    @prabhudaskamath13534 жыл бұрын

    Thanks Brandon..

  • @pratikedu8331

    @pratikedu8331

    3 жыл бұрын

    when the population mean is assumed to be 4 you mentioned there is a 0.16 probbablity of not rejecting the null hypothesis when it should have been rejected, my question is why should we have rejected the null hypothesis with a population mean of 4 when it falls in the brown region. anyone if understood my query please help me understand.

  • @MrDp297
    @MrDp2975 жыл бұрын

    How can u know the population standard deviation since all u have is a sample? Also shouldnt we use t statistic since the sample size is small?

  • @lamsonleung

    @lamsonleung

    5 жыл бұрын

    I hate stat

  • @vulnerablerummy
    @vulnerablerummy4 жыл бұрын

    04:19 just a small nitpick, but since we use the sample's parameter, shouldn't the standard deviation's symbol be 's' instead of sigma?

  • @saurabhaccentt

    @saurabhaccentt

    4 жыл бұрын

    I also noted that, but looks like he is assuming sample std as population std.

  • @shubhambasu642
    @shubhambasu6429 жыл бұрын

    You didn't considered the Zcritical value for the U=4, hence only the area between z=-2.33 and -1 should give the probability for type 2 error, which should come 0.15 or 15%. Since the area below that falls below the significance level for U=4, considering it should still support null hypothesis. In other words, only considering sample mean between 3.301 and 3.699 can lead to acceptance of null hypothesis when it should have been rejected leading to type 2 error. Please correct me if I am wrong !

  • @ricardoafonso7563
    @ricardoafonso75633 жыл бұрын

    . Thank you .

  • @pinksherbetsummer
    @pinksherbetsummer9 жыл бұрын

    Why is Z>+2.33 instead of -2.33? Most examples I see use Z

  • @pratikedu8331
    @pratikedu83313 жыл бұрын

    when the population mean is assumed to be 4 you mentioned there is a 0.16 probbablity of not rejecting the null hypothesis when it should have been rejected, my question is why should we have rejected the null hypothesis with a population mean of 4 when it falls in the brown region. anyone if understood my query please help me understand.

  • @francesmcbrien4785
    @francesmcbrien47852 жыл бұрын

    I'm 6 minutes into this. The last Starbucks example had 225 in the sample pop. This time it's 25. Will that change things?

  • @sunnypathak2335
    @sunnypathak23358 жыл бұрын

    I have confusion how to find out mean in this question.

  • @av5957
    @av59573 жыл бұрын

    I’m curious, does this beta have any relation to beta in linear regression ?

  • @BrandonFoltz

    @BrandonFoltz

    3 жыл бұрын

    Hi Andrew! It's is not related to beta weights in linear regression. Notation in statistics is all over the place. Makes it even more difficult when you see the same terms in different places!

  • @av5957

    @av5957

    3 жыл бұрын

    @@BrandonFoltz Brandon, thank you for clarifying this for me. For months, I’ve been trying to determine how they were connected but could never figure out how. I’m so glad I can put this to rest 👍🏼 Also, I have to say I am blown away by the work you put into your videos. You truly make it easy to understand the material. Thank you for all of your amazing work !

  • @ritvikshankar1057
    @ritvikshankar10573 жыл бұрын

    Isn't the z-statistic used when the population std dev is known? In the starbucks example the sample std dev is given - shouldn't the t-test be used?

  • @willluderman2522

    @willluderman2522

    2 жыл бұрын

    At 05:10 he states that sigma is known

  • @etikh404
    @etikh4044 жыл бұрын

    Nice narration but my confusion was the value 2.33 for Z. I got 1.66666

  • @nicholasng2115

    @nicholasng2115

    4 жыл бұрын

    (If this is incorrect, please correct me!)1.666 IS the value for z when using the sample mean of 3.5. I think the confusion is that, for what Brandon is demonstrating here, that test statistic and calculation is almost redundant. The case where z = 2.33 comes from when we set the area on the right tail as 0.01 (check the standard normal distribution tables). The fact that our test statistic value = 1.6666 = 3.699" as x-bar is 3.5. Hope that helps somewhat?

  • @francesmcbrien4785

    @francesmcbrien4785

    2 жыл бұрын

    @@nicholasng2115 What happened was he put 225 for n in the first example and in this one, he put 25 for n but kept the math from the first video example the same. How this old bird caught the error and no one else is beyond me.

  • @gooddeedsleadto7499
    @gooddeedsleadto74997 жыл бұрын

    Could you publish six sigma tutorials? thanks

  • @sharloable
    @sharloable8 жыл бұрын

    I am doing my masters in economics and your videos still give me great insights to things i already know. Thankyou for letting me know how flawed indian teachers are!!

  • @isabellaruso6282
    @isabellaruso6282Ай бұрын

    hi, i'm confused how you got 3.699

  • @manvendrasingh2243
    @manvendrasingh22439 ай бұрын

    Why use 2.33 and not 2.58?

  • @ruzannamanukyan1073
    @ruzannamanukyan10732 жыл бұрын

    if we take Alfa 0,01 our z in one tale should be 0,15% Why it is 2,33 which is correspond one tail Alfa 0,05 ?

  • @ruzannamanukyan1073

    @ruzannamanukyan1073

    2 жыл бұрын

    Sorry, I found in z table

  • @drkenny7928
    @drkenny79282 жыл бұрын

    Why dont you put this on udemy. It's so great to be free