Got a metrics exam in 4 days and your videos have genuinely saved me. Strides above the quality of teaching I have received, and incomparable to the online lectures we received towards the end of the semester. Just want to say huge thanks!
This series is a life-saving. THANK YOU, professor Lambert for this marvelous learning tool! Circumstances do not permit me to return to school, but I need this stuff urgently. It really means a lot. One question, though. You and I speak a different dialect of English and for the life of me I cannot understand what you called the y2t term that replaced the x2t term. "One of the ways serial correlation comes up is if we have a lacta?????dependent variable in our model"? Since a good part of this short video hinges on that one statement, I thought I would ask. Thanks again, for this course. Thank you. Thank you. Thank you.
Hi John, thanks for your message, and very kind words. My apologies for the confusion. I actually use the word, 'lagged' in the video. Thanks for pointing out this lack of clarity. All the best with your studies. Ben
Hi Ben, these videos are amazing, thank you so much! I was just wondering; if there is only serial correlation of order 1 but there is also endogenous regressors do you only need the standard t-test method once you have added all the independent variables to the regression on ehat, or do you still have to use the LM test? Thanks again for your help!
Hi Ben. Firstly, thanks so much for all your great videos! They are so amazingly helpful! I have a quick question - in this video, you are addressing how to test for SC when we have endogenous regressors - you say that a common example of this is when we have a lagged dependent variable. However, why is this endogenous? In an economic sense, I know 'endogenous' means 'determined within the model' but in econometrics, more specifically a variable Z is endogenous if E(u|Z) =/= 0. Say we have this model: Yt = a + bXt + cYt-1 + ut Why is it that Yt-1 is endogenous? I know that it is of course determined within the model as Yt-1 = a + bXt-1 + cYt-2 + ut-1, but why is it the case that E(ut|Yt-1) =/= 0 (i.e. endogenous in the econometric sense)? If this is the case, then surely we can't use OLS anyway as our estimates would be biased and inconsistent? I am just a bit confused here and was hoping you could clarify. Thanks again for making all these fantastic videos! Connor
Love u Ben can't think of anyone else I'd rather spend my last teenage evening with xx
I have my econometrics final tomorrow, this saved me soo much time thank you!
@@williamsingelstad3644 it's my turn now, wish me luck
@@somerandomguy3135 my turn now, wish me luck !! ahaha
@@meghna617 my turn now, wish me luck !! T_T
@@eskimogc4587 good luck!! you’ll smash it bestie
Got a metrics exam in 4 days and your videos have genuinely saved me. Strides above the quality of teaching I have received, and incomparable to the online lectures we received towards the end of the semester. Just want to say huge thanks!
This series is a life-saving. THANK YOU, professor Lambert for this marvelous learning tool! Circumstances do not permit me to return to school, but I need this stuff urgently. It really means a lot. One question, though. You and I speak a different dialect of English and for the life of me I cannot understand what you called the y2t term that replaced the x2t term. "One of the ways serial correlation comes up is if we have a lacta?????dependent variable in our model"? Since a good part of this short video hinges on that one statement, I thought I would ask. Thanks again, for this course. Thank you. Thank you. Thank you.
Hi John, thanks for your message, and very kind words. My apologies for the confusion. I actually use the word, 'lagged' in the video. Thanks for pointing out this lack of clarity. All the best with your studies. Ben
John Stro
What is the difference between this test and breush pagan test for checking heteroscedasticity?
Around the 3 minute mark, does anyone know why he skipped delta_1 and went from delta_0 straight to delta_2 in terms of notation?
Hi Ben, these videos are amazing, thank you so much! I was just wondering; if there is only serial correlation of order 1 but there is also endogenous regressors do you only need the standard t-test method once you have added all the independent variables to the regression on ehat, or do you still have to use the LM test?
Thanks again for your help!
Hi Ben. Firstly, thanks so much for all your great videos! They are so amazingly helpful! I have a quick question - in this video, you are addressing how to test for SC when we have endogenous regressors - you say that a common example of this is when we have a lagged dependent variable. However, why is this endogenous? In an economic sense, I know 'endogenous' means 'determined within the model' but in econometrics, more specifically a variable Z is endogenous if E(u|Z) =/= 0.
Say we have this model: Yt = a + bXt + cYt-1 + ut
Why is it that Yt-1 is endogenous? I know that it is of course determined within the model as Yt-1 = a + bXt-1 + cYt-2 + ut-1, but why is it the case that E(ut|Yt-1) =/= 0 (i.e. endogenous in the econometric sense)? If this is the case, then surely we can't use OLS anyway as our estimates would be biased and inconsistent?
I am just a bit confused here and was hoping you could clarify. Thanks again for making all these fantastic videos! Connor
3:32, you mean "corrected for endogeniety" right?
I was thinking the same thing haha. I'm glad I checked the comment section!
THANKYOU SO MUCH!!!! I Got an ecotrix presentation tomorrow, this saved me! :D
Well, this is a pleasant surprise :P
@ 3:37, I didn't clearly get when we include y(t-1) instead of independent variables (and not all of them, only x2(t)).
Veel te technisch he man
2:20
Português
dalbajobai