I would guess that a random sample of professional academic would get you a larger majority of people who is just in it for reserach, and maybe not too interested in teaching. All the while, a random sample of youtube turtoring accounts will produce a larger majority of people having the skill while simulatniusly being passioned about it. This discrepancy is not all that absurd when considering universities as only in second hand being all about teaching, and perhaps more of vassels of individual scientific attainment.
Dear Teacher! You saved me from dropping my economics class. Because of you I can now say that I can do Linear Regression analysis on Economical data. I searched everywhere for a simple explanation and you provided it. I am forever subscribed to your channel. Thanks a lot! If you need support sharing your videos let me know. You are the best!
Now I know that Econometrics itself is not too complicated to comprehend. It is the explanation method that matters. Thank you so much! I feel much more positive after watching just two of your videos.
I must say this is one of the best explanation about Linear regression, I started learning Machine Learning Last week it was kind of hard to grasp the concept but now it's like I was born to do this.
I convey my thanks to you. Its like life saving for me. It was too difficult to understand the class lecture regarding OLS regression Model. Now I could understand
Amazing indeed, particularly showing the population parameters and the sample parameters in the form of a simulation - it will stay in students' minds as long as they will work with samples. Thank you very much Sir!
I am here because of econometrics😭😭and i was finding it really tough and because of which I was tensed but somehow after watching your videos somehow my tension is relieved... Thank you
Thank you so much! Ur 18mins explanation is precise n better n easy to get the points than my 1.5hours lecture. I understand nothing from my professor haha
Thank you so much for all the content. I can say that you are better teacher the my Bachelor and MSc professors. You saved me from dropping my courses. Can you please upload videos about time series?
14:45 But thte p value isnt the probability that being the null hypotesis sustained, the probability of getting the observed values or more extreme ones (unlikely)?
but we are using coefficient 0.85 to set the intervals. what if we used different sample with different coefficient, would we get the same interval because the error is also gonna be different?
Apart from agreeing this is a very helpful video to many students, I would like to offer a correction to the rough explanation of the p-value, because it is unintentionally misleading for inexperienced people. *The p-value is NOT the probability there is no relationship*. It's very appealing to think the p-value can tell you this, because it sounds so understandable at first glance. Unfortunately, the problem is, you can't know the probability that there is no relationship. The correct way of interpreting the p-value (and this requires some careful reading): _Let's say there really is no relationship (which you can't know, but let's just assume), what is the probability that you will get the results you got in the sample of your study?_ I.e., how likely is it that your study could produce the results that it did, assuming there is no relationship in reality? That's the p-value. For instance, if your study produces sample results that really look like there is a relationship, and the p-value is 0.05, it means that if there happens to be no relationship in reality, you still have a 5% probability of getting those awesome sample results that suggest a relationship. However, when it's that unlikely to get those results if there is no relationship in reality, it's standard practice to assume the relationship in your study happened because there probably is a relationship in reality. Nevertheless, although the p-value may be very low, it's still possible there is no relationship in reality, and that your study just happened to have a fucked up sample of people (or plants, animals etc.) that randomly produced a data pattern that gave the impression of a relationship in your study. I only give this explanation to really interested students. Generally, however, I just say: "if the p-value is 0.05 or lower, you can assume you've PROBABLY found a relationship". I don't try to explain the p-value any further, because in my experience the concepts you need to understand it are very tricky to grasp for most inexperienced students. Here's a paper by Steven Goodman I found extremely helpful for understanding the misconceptions surrounding p-values: www.perfendo.org/docs/BayesProbability/twelvePvaluemisconceptions.pdf
Thank you for the comprehensive explanation of the p-value. I usually could not find any well-explained definition of p-value like this. Thank you again! it helps me a lot.
So, can I say that the p-value represents the probability of ending up with my particular sample (with which I built a regression model) from a population that in fact exhibits no such pattern as my model suggests? I’m a bit confused between whether you are referring to the probability of building a model the wrong way, or the probability of being given with the wrong data to begin with (=building a wrong model the right way). I’m not a native speaker of the English language, so I would much appreciate if you could be patient and add on to your already very elaborate explanation.
Thank you , grasping an intuitive explanation for p values for beginners is always tricky , I got a summary from khan academy that helped , the probability of the statistics assuming the null hypothesis is true, which is also the probability that determines whether you accept or reject the null hypothesis
Amazing video! Really liked the intuitive explanation of p-value (that its the probability that the null hypothesis is true). Just one suggestion - maybe at the end of the video, two disclaimers / caveats would be appropriate - (1) some rules of thumb need to be modified for one-tailed tests (in stead of the described two-tailed test) and (2) 5% level of significance is not considered sacrosanct by all
You can ... We have a null hypothesis and an alternate hypothesis so if for example if you reject your null hypothesis if your p value is below the 5% significance level(alpha = 0,05), then you accept the alternate hypothesis. Howeve, you can accept the null hypothesis if the p value is greater than alpha which means you will reject the alternate hypothesis. So one way or the other, rejecting one hypothesis means accepting the other.Hope it helped
He basically multiplied the std error by 2 and from the results of that answer, he added first and got the 751. Then, he subtracted that same number and got 954. Hope this helps.
why are free RUclips videos more helpful and easier to understand than my professor?
Life is full of surprises. 55555
Exactly what I am thinking
EXACTLY WHY :)
Agree.. Thanks for this info..
I would guess that a random sample of professional academic would get you a larger majority of people who is just in it for reserach, and maybe not too interested in teaching. All the while, a random sample of youtube turtoring accounts will produce a larger majority of people having the skill while simulatniusly being passioned about it. This discrepancy is not all that absurd when considering universities as only in second hand being all about teaching, and perhaps more of vassels of individual scientific attainment.
Dear Teacher! You saved me from dropping my economics class. Because of you I can now say that I can do Linear Regression analysis on Economical data. I searched everywhere for a simple explanation and you provided it. I am forever subscribed to your channel. Thanks a lot! If you need support sharing your videos let me know. You are the best!
i second this.
I third this.@@ankaah1230
X3
Now I know that Econometrics itself is not too complicated to comprehend. It is the explanation method that matters. Thank you so much! I feel much more positive after watching just two of your videos.
I must say this is one of the best explanation about Linear regression, I started learning Machine Learning Last week it was kind of hard to grasp the concept but now it's like I was born to do this.
It is amazing that some people are just so gifted at explaining things. Well, Explanined!!
Excellent video series! I encourage you to do more statistical topics (in English)...you are a gifted teacher.
He is a prof at Incae (one of the top business school in LATAM)
INDEED
you are the blessing for students i haven't seen such a comprehensive explanation ever in my life.God bless you.
I convey my thanks to you. Its like life saving for me. It was too difficult to understand the class lecture regarding OLS regression Model. Now I could understand
just mindblowing. Thank very very much for this clarification of P value and CI.
OMG, thank you for talking so practical. This video summarized one whole course of statistics I took.
You're the godfather of statistics! Thanks for providing these useful videos
Amazing indeed, particularly showing the population parameters and the sample parameters in the form of a simulation - it will stay in students' minds as long as they will work with samples. Thank you very much Sir!
Many many thanks! I can't express my gratitude in words. It was super helpful for me.
I cant yet definitely internalized everything but I am so happy of knowing this youtube channel... God bless...
I am here because of econometrics😭😭and i was finding it really tough and because of which I was tensed but somehow after watching your videos somehow my tension is relieved... Thank you
Why are you teaching like legend. Hats off..
Thank you so much. This is the best explanation I have seen so far. Thank you so much
Thank you so much! Ur 18mins explanation is precise n better n easy to get the points than my 1.5hours lecture. I understand nothing from my professor haha
Thank you very much for explaining statistics in simplest possible way. You are great teacher!
simply amazing. understood at 40 which was expected at 20
Pédagogiquement corret.
Right to the cross line. Thanks, you explanation is enough clear fro the mere mortal.
Great video and great example, you make it very intuitive and easy to understand!
Absolutely amazing explanation, Thank YOU professor.
Very well explained sir, The best video I ever came across
Excellent video with a clear and straight to the point explanation. This is really helpful! Keep up the good work and congratulations!
thank you now i can handle econometrics and found it very helpful. May God bless you.
Thank you so much for all the content. I can say that you are better teacher the my Bachelor and MSc professors. You saved me from dropping my courses. Can you please upload videos about time series?
After watching this video lecture it seems so simple. Loved the way you explained.
Excellent, logical, good examples, and easy explaination.
It is much clearer than my lecturer! Brilliant!
Wow...Cristal clear. Thanks 👍
I minute 10:23 of the video, I wish you explained the reason why you add and substract 2 x standard error on the slope...
Thank you so much. Your videos are awesome!
14:45 But thte p value isnt the probability that being the null hypotesis sustained, the probability of getting the observed values or more extreme ones (unlikely)?
Thank you so much for the excellent teaching stuff
best explanation by far
one among the best videos on statistics.
Mind-boggling 😀😀🔥
Your videos are fantastic, absolutely nailing the questions I have an explaining in a very thorough and understandable manner. Thanks!
i cannot thank you enough for this.
This is a wonderful video! Thank you, sir!
you are an amazing teacher!!
You are the best teacher everrrrr
Thank you sir! you are a great teacher
Amazing... You explained everything so well... love it
Thanks for making this video. It helped me a great deal understanding linear regression
Thanks for the video. Your explanation was very detailed and easily understood.
Wooow you are amazing! Nipped it by the bud! straight to the point and nice clear explanations!!!
Very well explained, a good addition to my lectures!
Very Excellent. Thanks
Can you provide a link to the Excel spreadsheet in 2:12?. Thanks
Marvelous. Great job done. The concepts are so finely described and are understandable in just one go.. Keep going.. *Thums Up*
You are awesome ❤
but we are using coefficient 0.85 to set the intervals. what if we used different sample with different coefficient, would we get the same interval because the error is also gonna be different?
You are a great Teacher
Thank you sir.
Wow,that was sooooo informative 👌
this is a very interesting lesson
Hi there, if you check at 13:44 , isnt that supposed to be 0.853 +- 2x0.051?
can anyone send the link to the course website? I would like this data the dataset to practice
Please provide link for dataset and thank you for your efforts.
@datamining where can I find the excel sheet you're using?
Amazing playlist!!
Sorry, What’s the course site that I can download files from ?
You are awesome 👌
Apart from agreeing this is a very helpful video to many students, I would like to offer a correction to the rough explanation of the p-value, because it is unintentionally misleading for inexperienced people. *The p-value is NOT the probability there is no relationship*. It's very appealing to think the p-value can tell you this, because it sounds so understandable at first glance. Unfortunately, the problem is, you can't know the probability that there is no relationship.
The correct way of interpreting the p-value (and this requires some careful reading):
_Let's say there really is no relationship (which you can't know, but let's just assume), what is the probability that you will get the results you got in the sample of your study?_ I.e., how likely is it that your study could produce the results that it did, assuming there is no relationship in reality? That's the p-value. For instance, if your study produces sample results that really look like there is a relationship, and the p-value is 0.05, it means that if there happens to be no relationship in reality, you still have a 5% probability of getting those awesome sample results that suggest a relationship. However, when it's that unlikely to get those results if there is no relationship in reality, it's standard practice to assume the relationship in your study happened because there probably is a relationship in reality. Nevertheless, although the p-value may be very low, it's still possible there is no relationship in reality, and that your study just happened to have a fucked up sample of people (or plants, animals etc.) that randomly produced a data pattern that gave the impression of a relationship in your study.
I only give this explanation to really interested students. Generally, however, I just say: "if the p-value is 0.05 or lower, you can assume you've PROBABLY found a relationship". I don't try to explain the p-value any further, because in my experience the concepts you need to understand it are very tricky to grasp for most inexperienced students.
Here's a paper by Steven Goodman I found extremely helpful for understanding the misconceptions surrounding p-values: www.perfendo.org/docs/BayesProbability/twelvePvaluemisconceptions.pdf
Thank you for the comprehensive explanation of the p-value. I usually could not find any well-explained definition of p-value like this. Thank you again! it helps me a lot.
So, can I say that the p-value represents the probability of ending up with my particular sample (with which I built a regression model) from a population that in fact exhibits no such pattern as my model suggests?
I’m a bit confused between whether you are referring to the probability of building a model the wrong way, or the probability of being given with the wrong data to begin with (=building a wrong model the right way).
I’m not a native speaker of the English language, so I would much appreciate if you could be patient and add on to your already very elaborate explanation.
Thank you , grasping an intuitive explanation for p values for beginners is always tricky , I got a summary from khan academy that helped , the probability of the statistics assuming the null hypothesis is true, which is also the probability that determines whether you accept or reject the null hypothesis
thank you so much, how can i get a not for this video?
just awesome! really good video!
Great work , I appreciate it !!
How do we fit the intercept to reduce error?
Thank you for helping me not fail 😭
AMAZING!!! Thank you!
God blessed you
Why does he highlight 1.96 standard deviation in graph instead of 2?
It's because the value 1.96 comes from the z table,since we've converted the distn into normal distn
Nicely described thanks :)
Great work!
thank you so much sir.
Really amazing...
Amazing video! Really liked the intuitive explanation of p-value (that its the probability that the null hypothesis is true). Just one suggestion - maybe at the end of the video, two disclaimers / caveats would be appropriate - (1) some rules of thumb need to be modified for one-tailed tests (in stead of the described two-tailed test) and (2) 5% level of significance is not considered sacrosanct by all
can you please provide the link to this worksheet which has the formula? Thank you!
Amazing video!
amazingly explained!
Hi... Great video.. appreciate if you let me know the reference book that you are using. Thnks
You teach better than these university teachers at yale :D
very good video
Good
thanks for the video
thank you so much!
well i failed statistics 2 terms in a row, now i feel like i can teach it...
is this the voice of Andrew ng?
very good, what about course website ?
A+ TEACHER
Nice explanation
I thought, according to my professor, we can never accept a hypothesis. We can only choose to reject or choose not reject, is this true?
You can ... We have a null hypothesis and an alternate hypothesis so if for example if you reject your null hypothesis if your p value is below the 5% significance level(alpha = 0,05), then you accept the alternate hypothesis. Howeve, you can accept the null hypothesis if the p value is greater than alpha which means you will reject the alternate hypothesis. So one way or the other, rejecting one hypothesis means accepting the other.Hope it helped
Thank You very Much =D
hi sir
i have a doubt in 1 part
0.853+- 2* 0.051 --> (0.751 , 0.954) where this range came from ?
He basically multiplied the std error by 2 and from the results of that answer, he added first and got the 751. Then, he subtracted that same number and got 954. Hope this helps.
Wheres the data set please
thank you Sir
best best best!!!!!!!!!!!!!
Make more videos pls.
HUGE