Видео

The best explanation of multinominal regression analysis

This is really helpful... You mentioned a handbook in the video. What is it, and is it possible to have access to it?

Congratulation!!!

Broo thank you so much, just done an entire question of my assignment using this one video😃😃

very helpful thank you

Great explanation! Thank you!

What happens when one of your interactions (Variable by LN_Variable) doesn't show up in the table? I put in 5 variables and their interactions with the log-odds, but only 4 show up. It also doesn't show up when I just add that one.

very helpful video, thank you.

Thank you so so much brother! Very helpful

Halo, I’m curious about the main effect of X no need to be consider?

Wow! Thank you! Found what I was looking for and very clear and easy to follow especially for an inexperienced user..

Thank you very much

First video I've seen on this that was actually really helpful! Thank you so much you are saving my dissertation!!!

👍

There is a lot of mouth noises and mouth clicks in this video.

This is so good. Thanks a lot for explaining it clearly. I hope you make such videos for other topics also

Flowing teaching

Hey thank you for your video. I have one question. If I have ordinal data (Likert scale) as my independent variables, should I put it into categorical covariates?

Thank you

what if theres 3 dependent variable outcomes? (3 stages of a disease)

This was helpfull, thank you!

Nice presentation

Great explanation, thanks!

Ma'am when I checking my assumption of outlier and applying CDF formula in that case it is showing "the expression ends unexpectedly" In this case, what can i do? I have 7 independent variables and 3 binary dependent variables.

very useful, thank you so much!

Thank you for a great video! I have a question: why is it important to centre the variables? Also, in my case, I want to look at the interaction effect of 'Incumbency' on the 'total amount of money' a candidate receives in an election when trying to predict 'vote share (%)' (DV). However, while money is a numerical value, I have coded 'incumbency' as a binary value of either 1 (they are an incumbent) or 0 (they are a challenger, or the seat is uncontested). Can I just multiply these together as you did in the video, even though my variables are different types of data? Thank you for your help!

If you included one variable at time, the OR would be a crude odds ratio... Therefore in this video the OR is actually adjusted. And variables are actually adjusting for the effect in each other

Excellent video!

Sir may i contact you for help

The best video about logistic regression that i have ever seen .. thank you for your effort

Please,sir could you help as by estimate the MMQR approach on stata ?

Thank you so much! Nice video 🤗

Is there a correlation between age and fare? Its only 0.09 whereas cut off is 0.7?

so the data is adjusted for survived group now?

A very big thanks to this very informative

Thank u for this video!!

how i can create the second graph at 2:29 ?

wonderful video! Thank you!!

excellent!

This is a great video, thank you! Can you recommend a similar analysis but using factor variables and not covariates?

this is amazing, thank you so much!! finally it makes sense :)

That was a really nice, clear example, and I love that you provided an example writeup as well. That is going to make things so much easier when analysing my data. Sincerely, thanks!

Please i do not get the. Stability for this, please help me out. line RGDPGR M2GR MPR LQR INFR BLR Year, legeng(size)medsmall]

I did it and everything went well. But one thing pokes mi in the eye. I've done the regression for 6 discrete variables as a possible predictor of a binary outcome. For one variable I got a really large number in the table, above 1000 for the Exp (B) and 15000+ for the Upper limit of CI interval 95%

If the overal p value is not significant that means that the variable shoudnot included in the model. Even if some of the variable modalities have a significant p

Amazing, thank you😊

Please, help! I have typed the 'boxtid' command but my Stata 14 refuses to recognise it.

Really nice video. Thank you! But i have a question, how do i do the Omnibus Test for a single variable, like, if i have 7 variabile and i want to start analysing the single omnibus test for single variables not by whole model. Hope you reply, thank you

Hi! thanks for the tutorial. I have one question tho: isn't the normality assumption of linear models supposed to be checked on the residuals of the model? i.e. on the pooled deviation between the values as predicted by the model and the actual data?

So helpful thank you!