Regression assumptions explained!

[@giovannifazio8546]

Why aren't my statistics classes at Uni like this one? Such a smooth visual and intuitive explanation. Best explanation ever, I'm so happy! Thank you!

[@GG-yk2nc]

i learned more about assumptions ( actually understand them, why we need them) in this video than my actual Graduate level stats class!!

[@ammar46]

Hey! If two variables are correlated we need to omit it, then wouldn't it cause omitted variable bias? When are we suppose to omit a variable?

[@freeandeasy9795]

I would have to say the same thing. Thanks for posting.

[@PunmasterSTP]

Hey, how have your studies been going over the past two years?

[@rajgill4432]

​@Ammar

[@rajgill4432]

​@@ammar46

[@xXxIMMORTALxXx]

It's really absurd that I've paid so much for my undergrad degree but the amount of knowledge received there isn't quite comparable to the vast amount of knowledge received on the internet and all these only at a fraction of a cost. I'm at this intersection in life where I'm considering whether I'd pursue a master's degree or continue working. The entry barriers such as GRE, GMAT score, the time preparing for those exams and the prices of taking them really make me think that instead of spending all that much time and money just to enter a university, it's a much better use of my time and money to delve into all the different advanced topics I can learn RIGHT NOW just by accessing the internet. I'm using your videos to try to get a new job during this pandemic. I don't know if I'd get it but regardless of the result I've learned so much about statistics from your videos already. Thanks a lot Justin! Keep doing what you do, you've helped a lot of people. I already subbed to your channel. Your channel is one of the best in building up statistical intuition and you really have that edge over other channels because you have an Australian accent. I find it very pleasant listening to stuff explained with an Australian accent haha.

[@zedstatistics]

Love this. You're inspiring me to make sure I keep it up! Thanks, Immortal!

[@PunmasterSTP]

Hey I'm just curious; what direction did you end up pursuing?

[@antygona-iq8ew]

Ha ha. I am topping my course content with KZbin tutorials as well. I actually spend my more time learning from youtube than from my dedicated learning platform.

[@xXxIMMORTALxXx]

@@PunmasterSTP Founding 2 startups simultaneously! One is an analytics consulting firm, another is a Metaverse. 😁 I got an undergrad degree in CS and I had careers both in software engineering and data analytics consulting. So, I find it natural for me to build both a consulting firm and a software house for the Metaverse. The past 2 years of my life I spent learning tons of stuff online have started to pay off. So very frickin glad I didn't end up taking GRE and being ripped off by some Uni somewhere in the US. 😂 Now I'm taking a bunch of Unity 3D courses from Udemy -- each costs me like $10. The unity skills will serve me so well in pursuing a startup in an industry that is expected to reach trillions in the coming years. My time in University was good guys but it was overprice in my opinion. I'm not against university enrollment at all, in fact the downside of learning online is it gets lonely sometimes. Had the university been cheaper, it would have been so much better. Oh and for those Europeans who might stop by -- yea yea I know it's free there, but I was an international student in Canada so... 😩

[@PunmasterSTP]

@@xXxIMMORTALxXx Oh snap that’s awesome! I hope your ventures continue to go well. If and when your software business puts out products, feel free to reply to this comment and I’ll gladly check them out!

[@wexwexexort]

This is so underrated. Should've get more views. Most of the people probably didn't stay for 47 mins. Poor people. They don't know what they missed.

[@recklessanil]

asin bayraklari

[@em70171]

This channel is an actual gem on learning statistics. Thank you so so much.

[@ABlueberryMuffin]

Where would I be without these videos! I am almost done with college and have survived by watching videos like these for studying! There is no way I would be where I am at without resources on youtube such as this. Thank you for making my many hours of studying more beneficial to me and making a clear understanding of the material. Awesome video

[@PunmasterSTP]

I'm assuming you finished college by now. If so, what are you up to currently?

[@ABlueberryMuffin]

@@PunmasterSTP unsure if my previous reply went through - but I am working in corporate finance at a major financial institution - working with forecasting, business strategy, and working closely with economics team to stress test various economic conditions. Studying this will pay off If your still in school make sure to do those internships and if you are near graduation apply for those rotational programs! Work on sending out those applications on a weekly/daily basis. Work hard on your linkedin :) things will work out for you and I wish u luck!

[@PunmasterSTP]

​@@ABlueberryMuffin Hey it's great to hear from you, and I am glad you are doing well! I actually chose a different career field, but I like to tutor students in a lot of different subjects. So from time to time I find something interesting to read up on and try to expand my horizons there. I also like replying to old comments and checking in with people to see who's out there. Cheers!

[@newgeneration8390]

I think you did a very good video. it shows you are a very perfect professor. I really enjoyed. love from Afghanistan

[@zedstatistics]

Back at you , pos-mis! Thanks for watching.

[@RossellaDeFalcoSoleyCelticHarp]

Your videos are brilliant - I almost can't believe they're all for free!! Thank you so much, I am a Ph.D. student, and I am using your vids as a regression refresher-to sum up everything before going into more advanced courses i.e. multilevel modelling :D thanks!

[@zedstatistics]

Tell your buddies :) glad you find them so useful!

[@PunmasterSTP]

How's your research going?

[@inytac]

I am a post graduate MBA student and i gotta say you’ve done a great job explaining this topic for me, Thank you

[@alexandragarcia9324]

This series saved me from finals. Cleared up 5 weeks of material from class. Thank you!

[@yannili3420]

Me too!!!! But mine is 14 weeks :X

[@PunmasterSTP]

@@yannili3420 Hey I know it's been years, but I just came across your comment and was curious. How'd your final end up going?

[@thesource8990]

I watch your videos every time I need to check if I have understood a standard textbook material adequately. Watching this after reading Dr. Gujarati's book. You are doing an amazing job. I can't thank you enough.

[@HA-vh3ti]

Unbelievable - how clear the flow of the lecture links the Topic, with examples to illustrate the solution to a Problem ! This is exactly how 'Net Global Knowledge' is increasing with such generous contribution from people like zedstatistics. Keep'em coming !

[@bahrimian]

best video series I came across in KZbin about statistics. thank you for the effort.

[@resap.9128]

I always go back to your videos once I need a review on stat. You are so good!

[@jaliu]

this channel is one of if not the best and most thorough only problem is not enough videos/topics!!!

[@shannawielinga8207]

This is amazing Zedstatistics, thank you so much. Also, your presentation is really properly build, with maximum learning value!

[@MrUndacovabrova]

I think the experiment you're talking about towards the end is Angrist (2000), where he used the conscription for the Vietnam War as a natural experiment to look at lifetime earning potential. I know this because I've gone through about 3 textbooks over the past few weeks and literally every single one of them mentions it lol. Either way this is the greatest educational series I've had the pleasure of encountering on KZbin and I'm almost certain to come back to it again and again throughout my future professional and academic career. I thoroughly enjoyed your little quips throughout, and your subtle sense of humor. Oh, and your Spotify playlist? Fuego. Thanks a lot Zed. You're a lifesaver. I hope you get Nobel consideration.

[@halimjun759]

It is amazing how you put assumptions in the last video of the regression series. My undergrad econometrics was just about assumptions and I just couldn't get it. Your video is so practical and I would like to show my appreciation for your video.

[@whetstoneguy6717]

I obtained a Masters in Economics from the University of Maryland. I wish I had watched your videos while I was taking the courses. This video was truly excellent. Thank you. Steven Greenspan

[@coolshoos]

It's hilarious how informative this video is... and all for free. Am I in grad school? Thanks a million, Justin.

[@sashpatra88]

Till now , though assumptions were covered in my class , it was pretty much theoretical stuff which I was least interested in. After watching yours video, now I could relate why are these assumptions so important for regression. Can't thank you enough for this crisp and to the point real explanation ...

[@rohitekka2674]

Months of panic comes to and end with your videos Justin. Thank you for this wonderful explanation. You've really organised your video well. If nothing else you've been a saviour to so many folks out there.

[@PunmasterSTP]

Regression assumptions explained? More like "Really, these videos are off the chain!" Thanks for putting together and sharing so much high-quality stuff.

[@wprowadzeniedostatystyki8328]

This is a very good educational video. So simple without annoying simplifications and hand-waving. Congrats!

[@loucuming6778]

Thanks

[@harryspeaks]

Just finished the whole series! Incredible work my friend, best stats class ever! I learn more in this series than in a whole year of stats class I swear! Big thanks!

[@Tweeteketje]

I love the structured approach of this video (issue, detection, remedy) and the explanation is so clear. This has helpe d me a lot!

[@shelbyf3671]

Thanks!

[@tylerwilson5769]

Yeah, I'd have to say you're the best teacher at my college. I'm getting my diploma from there but learned it all from KZbin.

[@potipiroon]

This is one the best explanations I have ever seen!

[@musclesmalone]

I cannot overstate how awesome this video is. Phenomenal! Thank you.

[@orthodontist777]

The video clarified all of my doubt within 1 hr. which was remain in for last 1 year. Thank you.

[@nguyenbaodung1603]

10:25 That's really the best residual plot I've ever heard. I love your videos. You're really my statistic hero!!

[@SuperSwaggaboiz]

I had the same realisation when he put it in that perspective!!

[@saurabhchoudhary4572]

Man your videos are like library on statistics, thanks for making these videos

[@NB19273]

Very clear and intuitive explanations! really helped me clear up understanding of regression. Thanks a lot!

[@philips2247540]

these series of videos were extremely helpful for my self-study. thank you

[@fame1437]

I wished I had known this channel a lot earlier. It would have been a great help for my studies.

[@SarveshwarInani]

It is very clear and effective. Great lesson ZEDSTATISTICS...Much Kudos.

[@theredviper24]

Can anyone explain to me how these regression assumption violations affect standard errors and coefficients? Everywhere I read about them they only mention that they become unreliable but no one tells why?

[@fatemehentezari2705]

Thank you for such useful videos. I guess I am learning all my graduate econometrics course from your videos.

[@naufalharits8259]

i have just finished this regression playlist and now i feel like i can conquer any questions on my regression exam next week 😂

[@kiranmahadeshwar9864]

Very very clear explanation. Easy to remember as all the assumptions combined into 6.

[@ryry8997]

really appreciate the effort you have gone to in constructing this series and your other vids mate, its helped digest alot of the stats material and clarified so many concepts. bloody ripper. liked and subbed. looking forward to more.

[@googledude5649]

Question - Omitted bias and the IV - In the end you talk about that you're cover that topic in more detail in another vid - can't find it and would love to hear your view on it ! - thanks a TON for your channel !

[@iamtrash288]

This is a great tutorial. It's so clear and well-put. Absolutely fantastic. Also love how in the examples he makes models that kinda seem to make sense at first glance and then shows that despite them being very nice in theory, there's some underlying problem. Very cool.

[@winniewu8581]

very good examples, I like the way he tells us what are the common misunderstandings, that helps a lot.

[@sharedknowledge162]

Danke!

[@kalyanakrishna6524]

Greatly appreciate your effort in putting up these videos. Apt examples and extreme clarity in explaining the concepts!! Hats off..

[@nicholaszulu7857]

This was very helpful and you explained the concepts in a simple and straight forward manner

[@valerieadachi9066]

Thank you for all the clear detail, care, and organization in your videos. if I pass stats it will be because of you!! ty!!

[@rodrigofreire2869]

Simply fantastic. Very weel detailed explanations. Congrats, Zed!

[@Artana2012]

excellent, I love all the examples, very clear, easy to understand, close to real life

[@aedi_idea]

Very nice video. Clear and informative. The example really hep in understanding the concepts.

[@shirok2460]

Great Tutorials!! Small comment on the autocorrelation: If I got it right, survey data for example can be correlated not only in time but mostly have spatial autocorrelation. (?)

[@keffje007]

This is all profesionally edited and well explained. Kudos!

[@naftalibendavid]

Nice work.

[@c00kie_monster37]

Can someone explain why including the age^2 term doesn't transform the equation from linear to quadratic at 11:44 ? I understand how the residual plot is now no longer curved, but including a squared parameter term (even if it is the same parameter in the prior term) HAS to make the equation quadratic and NOT linear, correct?

[@Ghnoom]

Zed, I love you man, like seriously! YOU are my gurdian angel in this jeorny I'm having. I love you, and I DON"T know how I would survive without you. Again, I f love you my dude!

[@bayarsaikhanbaasanbat9009]

Thank you so much for explaining all these concepts. I really needed to learn these as I’m about to do my graduate degree soon. You’ve covered everything I needed to know. Many thanks

[@mucheargaw6232]

hi thank you do send the procedure for merg files in spss thank you

[@h.i.sjoevall4213]

Thanks for the great videos! The quality is just excellent!

[@ShortVine]

44:02 you said here, you will make video on "omitted variable bias" but i checked in your channel there is no video related to this topic. can you please make 1 video on this topic. & thank you bro, you explained All Assumptions in a really simple way.

[@anothergirlinthecrowd5086]

four years late but yes please, been looking for one.

[@AdamuBig21]

From my heart, I say Thank You Sir. 🙏🏽

[@bhargavnits]

your videos are very helpful.. @11:07 : I have a question. By introducing age^2 along with age, we are getting a linear fit. But doesn't addition of age^2 feature, when there is age feature already , violate other assumption which is x variables themselves should not be related ?

[@gowezzmania880]

Great explanation and easy to digest for laypeople like me. I automatically subscribe to your channel and switch on the notification bell.

[@salmashaik1026]

Awesome!! Never understood all these concepts so clearly. Thank you!

[@JM-ty6uq]

At 12:00 you talked about turning the model into a multiple regression with an additional age^2 variable which adjusts the residual plot in the right way. Is this best practice to obtain better models?

[@bornhere13]

Brilliant! And with a plethora of examples. Well done.

[@MrReedMiester]

These videos are amazing! I wish I'd discovered them a long time ago

[@KnockoffProducer]

How about spatial autocorrelation? "Everything is related to everything else, but near things are more related than distant things." -- Tobler's first law of geography.

[@yannili3420]

I love the way you put it!!! You are right you've made these abstract things accordant to intuition

[@fagueyegueye5518]

Very helpful. Just amazing. Thanks for your good work.

[@ltoco4415]

Are these assumptions true for all types of regression i.e. Support Vector Regression, Lasso and Ridge regression, Stepwise regression etc or only for Linear Regression?

[@banashreeshiva4506]

Hi, Thank you for the videos! Sending you appreciations all the way from India :) I have a doubt here, if to solve multicollinearity we removed one of the variables, would it not cause omitted variable bias? For eg, if the number of people was removed in the example, then it would be the secret variable that is affecting both number of cars and motor accidents? You might say that some people might not travel or use public transport etc. But still, this might happen right?

[@vangelis9911]

Great job, big fan of yours!! But why is regression chapter so chaotic on your channel ?

[@omerfarukakbulut3665]

Finally i ended up here thanks for your effort on those videos for regression.

[@hirendesai5724]

I would like to understand the intuition behind calculation of standard error (formulae) next to slope coefficient. I haven't seen it in any of your videos or I have missed it

[@anupkumarmishra6363]

You are good. Thanks for the video, very clear explanations.

[@andriismykovskyi2999]

Hi Justin. You have mentioned that the transformation you'd performed in the linearity assumption part to better fit the model by adding squared age is not a p-hacking (14:04). Could you please explain why it is the case (or refer to some sources where I can read about it)?

[@jasonli4495]

extremely clearly explained! Thanks so much

[@microeconomia2561]

How about the assumption that "The variable x is not random and must take at least two different values"? What does it mean in regression analysis?

[@JuliusAu-ff4ip]

super clear explanation. Thank you !

[@existentialrap521]

Shoutout all the homies who made it this far. 🔥🔥🔥💴💯💯💯

[@BOHSociety]

Awesome work. Love it! Great refresher!

[@suhasshrinivasan]

Crisp and didactic. Wonderful! Thanks for the video. I do want to delve in a bit more detail on multi-colinearlity and interaction effects. I too thought that adding an interaction term would mitigate the issue of multi-colinearlity but it clearly won't. Any resources to better understand that? Thanks again!

[@zedstatistics]

I've got a multicollinearity video in the series and also a video on interaction terms :) Check my website.

[@ThamNguyen-ef4ez]

Thank you very much. your videos are really well-structured and help me a lot!!!

[@ayman1982100]

A question about the multicolearity, when you used the age and age square on a previous video, isn't it giving a multicolinearity??

[@michelmostoufi7707]

The same in the case of last example. If we use the Socio-economic status in regression model, does in not yield multicollinearity with years of education?

[@shubhrankavarmaable]

I think No... because both the terms -> age and age^2 are from column of one variable

[@BigTechWorldx2]

Not exactly. While age is monotonatic with age squared, they’re not linearly related to each other (they’re quadratically).

[@abhishekkrishnamurthy6074]

Can you please also include normal quantile plots in one of your videos?

[@eeshangautam]

for the multicollinearity part, if you consider your previous video with Interaction terms, there also you cannot keep one constant when you are changing otehr term whihc interacts with the one. RIGHT?

[@satishnanda2996]

Thank you so much for these ! Can you share the excel data for these examples to better understand the underlying concepts?

[@brittjane4486]

This is the best channel.

[@alwayz247]

Great video series. Can you create one regarding multicollinearity in linear regression?

[@zedstatistics]

Most likely will have it done by the end of the year. Stay tuned!

[@yoonkang2787]

Thank you so much! I really learned so much about regression though your videos (straightforward).

[@shahriarjami]

Great! Thank you so much for your clear explanation!

[@thalamus157]

This is really AMAZING! Thank you I Can’t Thank You Enough

[@wapogi8259]

Hello, Professor! Your video is direct and crisp, I should say. Am writing my dissertation now, and I would like to cite you and your video materials in my references. May I ask for more details like your full name and affiliation for attribution purposes? Also, would you find time to do a video lecture on Principal Components Regression (PCR) as well as a sequel to your lecture on Multi-Collinearity? Thank you very much, and more success and power... Romi Orio, PhD Candidate, Metro Manila, Philippines.

[@Anonymous-we4eu]

Hi, First of all thank you so much for your amazing videos!! I had a doubt though, why would introducing (age)^2 not violate linearity assumption? And can you please elaborate the justification you gave for it being a linear model - "y our lung function is linear as it relates to age and age^2" ?....i am so sorry for such a dumb question but it would be really great if you could help me out on this one! Thank you (Timestamp: 11:04)

[@preyanshugupta]

Awesome work. Just a small doubt. At around 10 min you mentioned that residual plot can be obtained by rotating the regression line to horizontal position. This is slightly incorrect as we do not take perpendicular distance between the regression line and the points for calculating residuals. Please correct me if my understanding is incorrect. Thanks.

[@zedstatistics]

Well it is true that the regression LINE rotates to the horizontal position. It is a visual aid for people who wonder where the residual plot comes from. Technically you are correct though, the 'dots' do not rotate around perfectly with the line for the reasons that you describe.

[@thesource8990]

Amazing effort! I can't thank you enough, yet, thank you very much! I have subscribed to your channel. This is the least I can do.

[@davosezra2143]

great explanation. so resourceful! thanks

[@kunalkochar9338]

@zedstatistics I learned a lot from your video, its very interactive and fun learning. I have a request: Can you please make a video on vapnik chervonenkis dimension?