I had to express my gratitude sir. I caught cancer late November had a surgery a week later and now cancer free. Late November was the last bit of my stats class. It's February 25, I was given an extension to take my final today. Your prerequisite list at around 3 1/2 minutes and several videos explaining those prerequisites just got me my degree. This was my last class and I haven't been in class III months and only had a review. I really only studied your videos for two days and your style made it possible.
Hi, I believe there's a typo in the df formula at 16:21. All the terms in brackets should be squared. That is, (s1^2/n1+s2^2/n2) should be (s1^2/n1+s2^2/n2)^2, (s1^2/n1) should be (s1^2/n1)^2, and (s2^2/n2) should be (s2^2/n2)^2. The degrees of freedom would then come out to be 55. However, this still would not be the most appropriate solution since this formula is Satterthwaite's approximation for unequal sample sizes. Key word: approximation. It is not a true generalization since it can be shown that for equal sample sizes, the degrees of freedom is exactly n1+n2-2=2(n-1). Therefore the degrees of freedom should be 2(30-1)=58. A very good video in all other respects. Cheers!
i was so overwhelmed and feeling like a complete failure, literally like i couldn't do anything right, university has affected my metal health so much so i really appreciate the beginning
One of the things mathematics education is getting across is the difference between conceptual understanding of mathematics, procedural understanding of mathematics and factual understanding of mathematics. There is a lot of emphasis in mathematics education on the latter two namely procedural understanding and factual understanding. It is very refreshing to find this video by Foltz Brandon which focuses exclusively on conceptual understanding. I join others in thanking him for producing such a video that emphasises conceptual understanding. Fundamental concepts are more than just basic concepts. How we see the world or think must not be brushed aside as just basic concepts. They are so fundamental that lack of understanding of key basic/fundamental concepts has led or is leading to challenges within mathematics understanding. May Foltz go on to produce more material on conceptual understanding. Thank you Foltz. Note to Reader: I have used the terms conceptual understanding of mathematics, procedural understanding of mathematics and factual understanding of mathematics as it is used by Daniel Willingham.
I have found your videos to be invaluable in helping me understand the concepts taught in my statistics class this term. The pace, content, graphics and style are ideal for learning. The encouraging intros are a nice touch also! Thank you!
Thank you for posting these videos...started watching them a couple of years ago...came back to it a few days ago and it all makes sense. Really love the way you take it step-by-step.
Thank you. I have gotten into a habit of watching the videos first before I even read the chapters. It helps me to understand the readings. Are your videos numbered for the progression in which they should be watched from start to finish? Any suggestions?
A very well explained and detailed methodology . Words about the fact that there is no such thing as negative minutes could be added + attention must be given to the typo found at 16:21, as pointed out by Emmanuel Burguete comments below.
dear Brandon thank you for this great series of lectures , I'm a physician and would like to learn statistics from A to Z especially statistics oriented towards research in medicine , so I have 2 questions please : 1- would you recommend me a playlist of yours where I can find what I'm specifically looking for . 2: is there a playlist of yours where the videos are organized in orderly way, so that I start by lecture 1 , then 2 etc ( I noticed that in the 34 videos playlist the video are not organised that way ) , thank you again for your really great lectures .
Dear Brandon , thank you for this great series of lectures. I recalculated the mean of call center 1 in R : 11.91, 11.36, 11.75, 12.14, 11.72, 11.61, 11.85, 12.16, 11.91, 12.12 which is not the same as the value you mention in the lecture = 11.853 as opposed to 11.91 you mentioned. I also recalculated the mean of the differences in R : -0.11 -0.66 -0.30 -0.04 -0.39 -0.46 -0.20 0.52 -0.48 0.47, which is not the same as you mention in the lecture = -0.165 as opposed to -0.11 you mentioned. Is there an explanation?
I guess he takes the statistics from the specific sample 1 and 2: which have x1 = 11.91 and x2 = 12.02, thus the difference is 0.11... But I am also not sure whether we should use the number from a specific sample or use the sample mean??
I am an addict with 15 years sober finally trying to get it all back together in life since losing to myself in high school. That is, I am in stats snd with a month to go I just slipped to a C+ from a B- where I'd been holding it for almost 3 months. That C+ is too close to a C- which is the end of the line, snd i'm 2 semesters from a bachelor's after years of overcoming addiction and then getting this far in education/ new life. Now that I am in a must win situation on the last 2 quizzes hearing the beginning of this tutorial got me past the idea that my professor really just wants me to fail. The independent t test or 2 populations is the next quiz and i'm alright with computations but the "different populations".... "sample the same equals reject".. "fail to reject" business leaves tremendous doubt and reacurring images of said professor and a red pen.
Your videos have helped me with stats more than anything else (even my teacher, even dare I say it..Khan). My exam is this coming Monday and I feel so much more confident. Thank you so much!!
In this video both the samples are of the same size. For independent samples t-test with samples sizes n1 and n2, what are the degrees of freedom? You gave a formula @ 16:21. But, I see it as this in my textbook: (n1+n2 - 2).
Many thanks for great lesson again. One question about the distribution plot at 28:22. I did not quite understand why value d1=-.11 is shown to the left of z=0 at the middle of the curve. To my understanding d1=-.11 should be right at the center because for our sample this is a d bar( blue colored at the picture) and it corresponds to z=0, but not some distance away from z=0 to the left. At the picture it looks like we are comparing two distributions, one with d bar=0 and the other with d bar -.11
Hi Brandon! I really do wonder why (at 26:20) you used sample mean difference value ( -0,11) instead of T-test value (-0,11 divided by group SEM 0,3507) for like in previous video with Z-test. So the brown arrow should be at -0,31. I think that is important difference.
Brilliant videos, indeed!. They wold be more interesting and easier to follow if you give links to related previous videos. I have been following the videos but got lost somewhere along the way, when trying to refer to a mentioned previous video. Could you please help to provide the links or do some sort of comprehensive cataloguing?. Many thanks.
Hi , I regularly watch your video to understand the basic concepts of statistics. you are just great. I understood these concepts of stat as you make the topic very simple and easy instead of making it a complicated . I have one question ---- is the standard deviation of the each individual sample for sample size n> 30 , same as the standard deviation of the sampling distribution of the sample means with same sample size . If Yes then is this true for sampling distribution of sample means with sample size n < 32 ( or n < 20), Please clarify.
Dear Brandon, I have followed your videos over the past year. You really are a rockstar teacher!!! I have one question about T-test. How do we differentiate between Independent t test and two-sample t? Thank you!!!!!!
Hello. Huge fan here. As you mention in the beginning of every video, I finally found a thing you could have done on the video. :) You could have mentioned the T statistic of the test.
Great video!! You are the bomb Brandon!! I am taking a second year stats courses and your videos have helped me improve my mark significantly from a 44 -91. Overall your videos are very easy to understand and you teach stats in a very intuitive manner. Keep up the great work!
I have a question. Why didn’t you calculate the t statistic of the sample to compare to the nonrejection area like you did in the previous video for the Two population z-test? By the way, I love your videos.
Really great videos. Very lucid way to make the concepts understand. It would be great if you can make videos of more advanced concepts of statistics, e.g. multivariate analysis, time series analysis, Bayesian statistics etc. I know i have made too many requests.
saptarshi mandal Thanks so much! I do have videos on Multiple and Logistic Regression. I have list a mile long that includes what you mentioned. I try to cover everything Stats 101 first before getting into more advanced topics. But I will never run out of topics to cover. Thanks for watching! :)
Hi Brandon. I am using a book as part of my University course titled Statistics by John McClave. In it the Degrees of Freedom formula is defined to be (n1 + n2 - 2).So for your example it would be = 58. But you have DoF of 28. Are you aware of the n1+n2-2 formula ?
Your videos are wonderful! I am now understanding the "why" which to me is much more important than the calculations. I am getting so stuck in my class- I am having some trouble finding the correct sequence of your videos- for instance I next need to go to related samples t tests and I also am getting confused with probability and sampling means. Is there an easier way for me to access these.
Hello. Excellent videos! May I ask at 10:30, step 4, when you mention (Theoretically) return to step #1 and repeat the process, it means that we will usually just be doing it based on the Xbar1 and Xbar2 from our collected data, since we will not have many chances to get many samples from our populations? In that case, we will only have one value of d. How do we get a distribution of differences with just one value of d? Thanks!
Hi Brandon, This is a t-test example, where the population's SD are not known. But you use SD from the previous video to calculate df and interval estimate. I stopped the video at that point to comment, and don't know if the rest of the video uses known population SD. Have I missed something?
Hi Brandon, I thumb you up... I watch some of your videos and they are very clear. Not sure whether you have a video to explain that "degree of freedom"... I'm sure this degree of freedom is not intuitive for most of us.
hi thanks for the really helping lectures, i have a Q , u took the d bar which is the mean of differences but when u apply it to the formula u applied the difference between 2 means not the mean of the differences which i calculated from the table you showed to be equal to -0.157 can u please explain that thank u
Hi Brandon. When given 2 different sample means, 2 different sigmas, and 2 different sample sizes: isn't there an excel function that can compute the calculation?
Hello! Excel has both the two population z- and t- tests. For the t-test I only think it has "assume unequal variances" but does not allow you to specify each one. The z-test however does have inputs for specific values. This is in the Data Analysis ToolPak addin by the way. - B
Great video. :) Thank you!! One question: 19 minutes in, you get the t-critical = 2.048. Could be I don't understand something, but from my T-table I get t-critical =1.701. What is right here?
Thanks! You are looking at the a=.05 in one tail of the distribution. This would mean a=.10 for both tails. In this case you have to split the a=.05 across both tails, so a=.025 in upper and lower. Find a=.025 with df=28 and you will land on 2.048 :) Hope that helps!
Hi. Would you by any chance still have the Excel of the data set you used for this video? I kinda got into the habit of computing for myself after watching your video. I know this was 3 years ago but asking anyway :)
Could someone please explain: 13:00 how does the formula f or the "Standard Error of the Difference" Sx1-x2 come about? This is the standard deviation of the *difference* between sample means, right?
I'm trying to solve an exercise in which I'm not given a mean, but I'm told that they want to test a new drug, so 80 people get tablets with the drug and 120 get a placebo. 60 of those who got the drug have a positive effect, while in the other group 50 get a positive effect. How do I calculate whether the drug has a statistically relevant effect or not?
Just like what others have asked (and left unanswered), why didn't you just calculate the t-statistic then see if we are to reject the null hypothesis or not, just like what you did in the previous video about two population z-test? That's what I find confusing about this video. :(
@@ok9621 If you know the population standard deviation you use the z-test, if you don't know the population standard deviation then you use the t-test. In t-test you calculate the degree of freedom.
Hello, can somebody tell me If it is appropriate to calculate T-test statistic with s1(sample mean 1) and s2(sample mean 2) as we did this in prev video for Z-test statistic (sigma is known)? Thanks in advance!
In the call center example you provide 10 sample measurements at 10:10..Is this not the same sample used for the later calculations where n=30?. Would the calculation of the sample standard deviation shown at 10:10 yield .25? How can we calculate sample standard deviation s from measurements taken from a sample?
Brad Mello Hi! I am not sure if I understand your question completely but I'll try to clear things up. At 11:00 what we are seeing is 10 samples each of size 30. This is the sampling distribution of differences just for a conceptual foundation. The call center problems later in the video are not directly related to 11:00. They have their own data I just give you. And finally the sample standard deviation s, is a common calculation used to estimate the population standard deviation using a sample.
***** Thanks for clearing that up. I'm learning following your statistics videos and directly applying the knowledge in JMP software package. Do you think a paired t-test will be a useful technique to compare two sample populations to test if they have similarities?
Hello...why the question in 17:43 is using t distrubution?according to central limit theorem when sample size bigger or equal to 30...we can assume tge sample is normally distrubuted....and we can use z right?
Hello! Technically when the standard deviation is unknown the t-distribution should be used. At n = 100 the t- and z- are nearly identical so they are the "same" in practice. I always advocate using the t-dist whenever sigma is unknown because it is the more conservative approach.
Can someone please help? The example used in this video took large number of samples from the population (multiple times of 30 samples) to form sampling distribution so that we can get the variance of the sampling distribution. However, often times, we do not have the luxury of forming this sampling distribution. How can we get the variance s1 and s2 in this case when we only draw one group of sample to compare their mean? Possibly using pooled variance assuming they are equal? Thank you in advance everyone!
Can anyone please explain the following point In the previous video when the population std deviation if known he took the population mean difference (D) which was 0.Whereas in this video although the population mean for each population was given why did he take the difference between two sample mean to calculate D?
I had to express my gratitude sir. I caught cancer late November had a surgery a week later and now cancer free. Late November was the last bit of my stats class. It's February 25, I was given an extension to take my final today.
Your prerequisite list at around 3 1/2 minutes and several videos explaining those prerequisites just got me my degree. This was my last class and I haven't been in class III months and only had a review. I really only studied your videos for two days and your style made it possible.
hope you are doing well
So happy for you on both milestones, I am happy you are healthy and that you got your degree as well!
wish you happy time, sir.
Get well soon
.
Hope you are well
The wonders of the kindness of so many who recorded their work on these YT opportunities
The world is good
.
Hi, I believe there's a typo in the df formula at 16:21. All the terms in brackets should be squared. That is, (s1^2/n1+s2^2/n2) should be (s1^2/n1+s2^2/n2)^2, (s1^2/n1) should be (s1^2/n1)^2, and (s2^2/n2) should be (s2^2/n2)^2. The degrees of freedom would then come out to be 55.
However, this still would not be the most appropriate solution since this formula is Satterthwaite's approximation for unequal sample sizes. Key word: approximation. It is not a true generalization since it can be shown that for equal sample sizes, the degrees of freedom is exactly n1+n2-2=2(n-1). Therefore the degrees of freedom should be 2(30-1)=58.
A very good video in all other respects. Cheers!
Emmanuel Burguete thank you so so so much!!
The typo correction herein suggested applies.
i was so overwhelmed and feeling like a complete failure, literally like i couldn't do anything right, university has affected my metal health so much so i really appreciate the beginning
You just brought me back to mathematics. Very nice, understandable, and easy way to teach. Truly appreciated.
That pep talk at the beginning tho! *tears*
I don't think he has any idea how much I needed that pep talk!
THIS DUDE IS LIFE
LOL
One of the things mathematics education is getting across is the difference between conceptual understanding of mathematics, procedural understanding of mathematics and factual understanding of mathematics. There is a lot of emphasis in mathematics education on the latter two namely procedural understanding and factual understanding. It is very refreshing to find this video by Foltz Brandon which focuses exclusively on conceptual understanding. I join others in thanking him for producing such a video that emphasises conceptual understanding. Fundamental concepts are more than just basic concepts. How we see the world or think must not be brushed aside as just basic concepts. They are so fundamental that lack of understanding of key basic/fundamental concepts has led or is leading to challenges within mathematics understanding. May Foltz go on to produce more material on conceptual understanding. Thank you Foltz. Note to Reader: I have used the terms conceptual understanding of mathematics, procedural understanding of mathematics and factual understanding of mathematics as it is used by Daniel Willingham.
I have found your videos to be invaluable in helping me understand the concepts taught in my statistics class this term. The pace, content, graphics and style are ideal for learning. The encouraging intros are a nice touch also! Thank you!
Thank you for posting these videos...started watching them a couple of years ago...came back to it a few days ago and it all makes sense. Really love the way you take it step-by-step.
Thank you. I have gotten into a habit of watching the videos first before I even read the chapters. It helps me to understand the readings. Are your videos numbered for the progression in which they should be watched from start to finish? Any suggestions?
me too. He provides good visual description. Books make it so abstract.
A very well explained and detailed methodology . Words about the fact that there is no such thing as negative minutes could be added + attention must be given to the typo found at 16:21, as pointed out by Emmanuel Burguete comments below.
I so needed to hear your introduction. Thank you for your compassion.
Your intro message is the greatest I have heard and I am just in love with your videos!
dear Brandon
thank you for this great series of lectures , I'm a physician and would like to learn statistics from A to Z especially statistics oriented towards research in medicine , so I have 2 questions please : 1- would you recommend me a playlist of yours where I can find what I'm specifically looking for . 2: is there a playlist of yours where the videos are organized in orderly way, so that I start by lecture 1 , then 2 etc ( I noticed that in the 34 videos playlist the video are not organised that way ) , thank you again for your really great lectures .
thanks for the video! A question tho: from other place i saw df=n1+n2-2, why it's so complicated here, and it's only half of n1+n2-2? Thank you.
Dear Brandon , thank you for this great series of lectures.
I recalculated the mean of call center 1 in R : 11.91, 11.36, 11.75, 12.14, 11.72, 11.61, 11.85, 12.16, 11.91, 12.12 which is not the same as the value you mention in the lecture = 11.853 as opposed to 11.91 you mentioned.
I also recalculated the mean of the differences in R : -0.11 -0.66 -0.30 -0.04 -0.39 -0.46 -0.20 0.52 -0.48 0.47, which is not the same as you mention in the lecture = -0.165 as opposed to -0.11 you mentioned.
Is there an explanation?
I guess he takes the statistics from the specific sample 1 and 2: which have x1 = 11.91 and x2 = 12.02, thus the difference is 0.11... But I am also not sure whether we should use the number from a specific sample or use the sample mean??
Millions of thanks for your clear and vivid explanations it helps me a lot. Two thumbs up.
I am an addict with 15 years sober finally trying to get it all back together in life since losing to myself in high school. That is, I am in stats snd with a month to go I just slipped to a C+ from a B- where I'd been holding it for almost 3 months. That C+ is too close to a C- which is the end of the line, snd i'm 2 semesters from a bachelor's after years of overcoming addiction and then getting this far in education/ new life. Now that I am in a must win situation on the last 2 quizzes hearing the beginning of this tutorial got me past the idea that my professor really just wants
me to fail. The independent t test or 2 populations is the next quiz and i'm alright with computations but the "different populations".... "sample the same equals reject".. "fail to reject" business leaves tremendous doubt and reacurring images of said professor and a red pen.
Your videos have helped me with stats more than anything else (even my teacher, even dare I say it..Khan). My exam is this coming Monday and I feel so much more confident. Thank you so much!!
In this video both the samples are of the same size.
For independent samples t-test with samples sizes n1 and n2, what are the degrees of freedom? You gave a formula @ 16:21.
But, I see it as this in my textbook: (n1+n2 - 2).
Many thanks for great lesson again. One question about the distribution plot at 28:22. I did not quite understand why value d1=-.11 is shown to the left of z=0 at the middle of the curve. To my understanding d1=-.11 should be right at the center because for our sample this is a d bar( blue colored at the picture) and it corresponds to z=0, but not some distance away from z=0 to the left. At the picture it looks like we are comparing two distributions, one with d bar=0 and the other with d bar -.11
Hi Brandon!
I really do wonder why (at 26:20) you used sample mean difference value ( -0,11) instead of T-test value (-0,11 divided by group SEM 0,3507) for like in previous video with Z-test. So the brown arrow should be at -0,31. I think that is important difference.
You are definitely an awakener, you awaken the confidence that I needed. Pep talk=A+
Thank you for your videos. I have been all over UA-cam and yours are by far the easiest to understand. I have used your videos all semester!
Brilliant videos, indeed!. They wold be more interesting and easier to follow if you give links to related previous videos. I have been following the videos but got lost somewhere along the way, when trying to refer to a mentioned previous video. Could you please help to provide the links or do some sort of comprehensive cataloguing?. Many thanks.
I am passing stats due to these vids. Thank you. I was so confused with this online course before finding these. Thank you so much!
Excellent video. I have been having trouble calculating t-tests when comparing two populations. The tutorials are great.
it would really be helpful if you posted links with practice problems. Overall, Great job on the videos! Extremely helpful!
Hi , I regularly watch your video to understand the basic concepts of statistics. you are just great. I understood these concepts of stat as you make the topic very simple and easy instead of making it a complicated . I have one question ---- is the standard deviation of the each individual sample for sample size n> 30 , same as the standard deviation of the sampling distribution of the sample means with same sample size . If Yes then is this true for sampling distribution of sample means with sample size n < 32 ( or n < 20), Please clarify.
Dear Brandon, I have followed your videos over the past year. You really are a rockstar teacher!!! I have one question about T-test. How do we differentiate between Independent t test and two-sample t? Thank you!!!!!!
Hello. Huge fan here. As you mention in the beginning of every video, I finally found a thing you could have done on the video. :) You could have mentioned the T statistic of the test.
Great video!! You are the bomb Brandon!!
I am taking a second year stats courses and your videos have helped me improve my mark significantly from a 44 -91. Overall your videos are very easy to understand and you teach stats in a very intuitive manner.
Keep up the great work!
Yet another statistics class your videos have helped me through (Statistical Reasoning in Psychology)!
Love the tutorial, very helpful for me.
I have a question. Why didn’t you calculate the t statistic of the sample to compare to the nonrejection area like you did in the previous video for the Two population z-test? By the way, I love your videos.
Best Statistics explanations on you tube. Saved my ass.
Really great videos. Very lucid way to make the concepts understand. It would be great if you can make videos of more advanced concepts of statistics, e.g. multivariate analysis, time series analysis, Bayesian statistics etc. I know i have made too many requests.
saptarshi mandal Thanks so much! I do have videos on Multiple and Logistic Regression. I have list a mile long that includes what you mentioned. I try to cover everything Stats 101 first before getting into more advanced topics. But I will never run out of topics to cover. Thanks for watching! :)
***** I have also watched your logistic regression videos. They are also very great indeed. Thanks again.
Hi Brandon. I am using a book as part of my University course titled Statistics by John McClave. In it the Degrees of Freedom formula is defined to be (n1 + n2 - 2).So for your example it would be = 58. But you have DoF of 28. Are you aware of the n1+n2-2 formula ?
Beautifully explained. Best of all Stats help.
hi. great video!. I got a question for you: the statistical analysis is the same for means less than 1?
Your videos are wonderful! I am now understanding the "why" which to me is much more important than the calculations. I am getting so stuck in my class- I am having some trouble finding the correct sequence of your videos- for instance I next need to go to related samples t tests and I also am getting confused with probability and sampling means. Is there an easier way for me to access these.
I donated. I appreciate your videos.
Great video! Can you also add critical thinking of null and alternative hypotheses?
Thanks for everything. Is there a way to get the powerpoint (pdf) of your lectures?
Hello. Excellent videos!
May I ask at 10:30, step 4, when you mention (Theoretically) return to step #1 and repeat the process, it means that we will usually just be doing it based on the Xbar1 and Xbar2 from our collected data, since we will not have many chances to get many samples from our populations?
In that case, we will only have one value of d. How do we get a distribution of differences with just one value of d?
Thanks!
i think that is what the confidence interval for.
Awesome mr. Foltz. Very much enjoyed your video. Greetings from Holland.
Edit: AND you're into Impreza's. It doesn't get better than this
Thanks Brandon. Question: why we did not use t_score method for this (similar to previous video where we used z score method) ?
Hi thanks for the great videos. Don't you think that the sample size n1 and n2 at 17.24 should be 10 and not 30?
Hi Brandon,
This is a t-test example, where the population's SD are not known. But you use SD from the previous video to calculate df and interval estimate. I stopped the video at that point to comment, and don't know if the rest of the video uses known population SD. Have I missed something?
Hi Brandon, I thumb you up... I watch some of your videos and they are very clear. Not sure whether you have a video to explain that "degree of freedom"... I'm sure this degree of freedom is not intuitive for most of us.
thank you for the positive message in the beginning, it made me feel better :3
Thanks so much for all your videos. So helpful and inspiring. Gracias.
Thanks for this sir... keep uploading contents that helps
Thanks Brandon for helping, I wish my class lectures were as good as you.. BTW I am from Bangalore!!!
Brandon, quick question, when you are not given standard deviation for both population, is there a particular change in calculations??
hi thanks for the really helping lectures, i have a Q , u took the d bar which is the mean of differences but when u apply it to the formula u applied the difference between 2 means not the mean of the differences which i calculated from the table you showed to be equal to -0.157 can u please explain that
thank u
Hi Brandon.
When given 2 different sample means, 2 different sigmas, and 2 different sample sizes: isn't there an excel function that can compute the calculation?
Hello! Excel has both the two population z- and t- tests. For the t-test I only think it has "assume unequal variances" but does not allow you to specify each one. The z-test however does have inputs for specific values. This is in the Data Analysis ToolPak addin by the way. - B
Very clear explanation
Great video. :) Thank you!!
One question: 19 minutes in, you get the t-critical = 2.048. Could be I don't understand something, but from my T-table I get t-critical =1.701. What is right here?
Thanks! You are looking at the a=.05 in one tail of the distribution. This would mean a=.10 for both tails. In this case you have to split the a=.05 across both tails, so a=.025 in upper and lower. Find a=.025 with df=28 and you will land on 2.048 :) Hope that helps!
Aha! Got it. thanks!! :D
I like your Videos, salute from India.
can we use the 2-sample independent sample t-test with unequal variance for comparing two 85th percentile values similar to mean values?
Hi. Would you by any chance still have the Excel of the data set you used for this video? I kinda got into the habit of computing for myself after watching your video. I know this was 3 years ago but asking anyway :)
Thank you so much! Thanks to you I wont be failing STAT !!!
You are a LIFE SAVER!! Thank you!
Could someone please explain: 13:00 how does the formula f or the "Standard Error of the Difference" Sx1-x2 come about? This is the standard deviation of the *difference* between sample means, right?
This is ridiculously awesome. Thank you.
13:03 The standard error of d - can it be calculated from d mean and di (the last column in 10:48)?
Is it wrong to calculate degrees of freedom by n-2 in this case?
I'm trying to solve an exercise in which I'm not given a mean, but I'm told that they want to test a new drug, so 80 people get tablets with the drug and 120 get a placebo. 60 of those who got the drug have a positive effect, while in the other group 50 get a positive effect. How do I calculate whether the drug has a statistically relevant effect or not?
Could you please make the slides available as a .pdf?
Thanks for the great insight.
Thanks for the pep talk! :D
How did you get the first and second standard deviations? 7:52
Excellent as usual
Your vote of confidence is wonderful. I have my star final tomorrow
Loved this lesson. So helpful!
Great videos! Really enjoying the content! Thanks!
How can determine which tail to use when the claim is that a population's mean is bigger than another population's mean?
Just like what others have asked (and left unanswered), why didn't you just calculate the t-statistic then see if we are to reject the null hypothesis or not, just like what you did in the previous video about two population z-test? That's what I find confusing about this video. :(
Thanks for making this video. Now I learned what the mean difference is. I've watched many of your vids so far. I'm becoming quite a crack ;)
I have finals tomorrow and don't even know when I should use t test or z test help!
me right now except it's technically today
@@ok9621 If you know the population standard deviation you use the z-test, if you don't know the population standard deviation then you use the t-test. In t-test you calculate the degree of freedom.
Thanks, Quite useful presentation....
Hello, can somebody tell me If it is appropriate to calculate T-test statistic with s1(sample mean 1) and s2(sample mean 2) as we did this in prev video for Z-test statistic (sigma is known)? Thanks in advance!
In the call center example you provide 10 sample measurements at 10:10..Is this not the same sample used for the later calculations where n=30?. Would the calculation of the sample standard deviation shown at 10:10 yield .25?
How can we calculate sample standard deviation s from measurements taken from a sample?
Brad Mello Hi! I am not sure if I understand your question completely but I'll try to clear things up. At 11:00 what we are seeing is 10 samples each of size 30. This is the sampling distribution of differences just for a conceptual foundation. The call center problems later in the video are not directly related to 11:00. They have their own data I just give you. And finally the sample standard deviation s, is a common calculation used to estimate the population standard deviation using a sample.
***** Thanks for clearing that up. I'm learning following your statistics videos and directly applying the knowledge in JMP software package. Do you think a paired t-test will be a useful technique to compare two sample populations to test if they have similarities?
Paired t-test is only appropriate when the measurements are from the same person or thing. Otherwise use standard unpaired t-test.
Hello...why the question in 17:43 is using t distrubution?according to central limit theorem when sample size bigger or equal to 30...we can assume tge sample is normally distrubuted....and we can use z right?
Hello! Technically when the standard deviation is unknown the t-distribution should be used. At n = 100 the t- and z- are nearly identical so they are the "same" in practice. I always advocate using the t-dist whenever sigma is unknown because it is the more conservative approach.
oh...tq:)
Maaaaate you re such a legend! thanks for your effort ;)
You are the best dude!
awesome lesson
Can someone please help? The example used in this video took large number of samples from the population (multiple times of 30 samples) to form sampling distribution so that we can get the variance of the sampling distribution. However, often times, we do not have the luxury of forming this sampling distribution. How can we get the variance s1 and s2 in this case when we only draw one group of sample to compare their mean? Possibly using pooled variance assuming they are equal? Thank you in advance everyone!
Hi can you please provide transcript of this video.. thank you would be a great help...
Thank you very very much
why not use t statistic to reject/ accept the null hypothesis in this video?
How to calculate sample size requirement for this? Where to get the standard errors in order to calculate sample size requirement? Nothing makes sense
Great video - thanks!
Great works keep it up
Thank you.
Hi Brandon, what a nice introduction. Thanks for the motivation and talking about the chance to connect, that cool! Thank you. Luca
Can anyone please explain the following point
In the previous video when the population std deviation if known he took the population mean difference (D) which was 0.Whereas in this video although the population mean for each population was given why did he take the difference between two sample mean to calculate D?
i have the same question. why not calculate the T-statistic directly?
sorry im a bit lost. why is there no t statistics for this example
thanks fultz
Is this whole process the t test?