Awesome explanation! But I am still wondering if you have something on how to use mixed effect models for Censored dependent variable (e.g. percentage data ranging from 0-100%). Thank you!
Hi I was wondering if you could use the same procedure when the observations are nested within two levels that are not related, that is Level 2 is not nested within Level 3. For example observation in nested within "State" and observation is also nested within "Person"
The analogy doesn't translate very well literally though my hand calculation is quite close to the -xtmixed-result. The reason is that -xtmixed- is fitting the entire model using all of the observations and variables simultaneously. You won't get the identical result if you consider one part of the model in isolation. The results also depend on the estimation method used such as ML or REML.
Hi, I am working on multiple time series regression model. I check through corrgram, pperron and dfuller tests [in stata] that the dependent variable is stationary at level 3 while independent variables are stationary at level 2 or 3. Can we still run VECM when the variables are cointegrated but stationary at different levels? An other question is that VAR and VECM are the candidate estimation techniques for multiple regression time series. Is there any other technique for multiple time series regression?
All, I'm using Stata 13. Does anyone know how to group-mean center the variables? grand mean centering worked with "mcenter". In Stata 14, group mean centering is just "center", but I'm on a different version than a person working wtih me.
What would you do if you have two bases for clustering, but not on two levels? From your example, let say region and industry; so cases are clustered in regions and in industries?
Is it possible to apply Multilevel models when data set consists of 3 IVs of categorical nature and 1 or more continuous DVs ?.......... My data set consists of 3 categorical IVs (1st IV consists of 2 categories, Second IV also consists of 2 Categories and 3rd IV consists of 3 categories).
The *xtmixed* command has been renamed to *mixed* . You can fit a model with *mixed* and then use the *collect get* command to collect the results from the command. You would use a named expression with *collect get* to display the random effects as they are shown in the output from *mixed* . Then you would simply lay out your table with *collect layout* . Please email us at tech-support@stata.com if you would like to see an example.
Sir, I am currently working on a meta analysis. But I am not able to produce association studies of allele contrast (A vs a). Except that all are done but can not understand how to calculate A vs a frequency. If you help me out it will help me a lot. Thanks
I was so nervous about using Stata for HLM, but this was immensely helpful and really comforting. Thank you for the great video!
Very nicely done. These are a terrific resource. Thank you, Chuck and Stata!
Excellent video, clear explanation, smooth lecture. Just awesome. You should do more videos!
Chuck is a genius for teach, genius for understand the concepts!!!!
So many thanks for your videos. I am doing biostatistics and I can testify that your video are for a great help to me. Glod bless you and stata
Chuck Huber! you are awesome! keep producing videos please.
Very helpful!! Expecting part two.
Excellent video, very intuitive graphics, thanks a ton
Note that the "xtmixed" command has been superceded by the "mixed" command in newer versions of stata
awesome! very clear explanation.
Thanks for this great video!
Awesome! Keep up the great work!
Excellent video sir.
You can download the dataset used in this example by typing: use www.stata-press.com/data/r13/productivity.dta
Awesome explanation! But I am still wondering if you have something on how to use mixed effect models for Censored dependent variable (e.g. percentage data ranging from 0-100%). Thank you!
Very useful - thanks!
Awesome !
Excellent
Thanks for this. It'd be great to have a video on multilevel modelling with gsem (GUI + code).
Hi I was wondering if you could use the same procedure when the observations are nested within two levels that are not related, that is Level 2 is not nested within Level 3. For example observation in nested within "State" and observation is also nested within "Person"
Stata 12, is the option 'covariance(independent)' default? In Stata 14, how can adopt this opt command?
The analogy doesn't translate very well literally though my hand calculation is quite close to the -xtmixed-result. The reason is that -xtmixed- is fitting the entire model using all of the observations and variables simultaneously. You won't get the identical result if you consider one part of the model in isolation. The results also depend on the estimation method used such as ML or REML.
thnx a lot
Hi, I am working on multiple time series regression model. I check through corrgram, pperron and dfuller tests [in stata] that the dependent variable is stationary at level 3 while independent variables are stationary at level 2 or 3. Can we still run VECM when the variables are cointegrated but stationary at different levels?
An other question is that VAR and VECM are the candidate estimation techniques for multiple regression time series. Is there any other technique for multiple time series regression?
Please contact technical support with your question at www.stata.com/support/tech-support/contact.
All,
I'm using Stata 13. Does anyone know how to group-mean center the variables? grand mean centering worked with "mcenter". In Stata 14, group mean centering is just "center", but I'm on a different version than a person working wtih me.
What would you do if you have two bases for clustering, but not on two levels? From your example, let say region and industry; so cases are clustered in regions and in industries?
u run two different models, one for each cluster
Is it possible to apply Multilevel models when data set consists of 3 IVs of categorical nature and 1 or more continuous DVs ?..........
My data set consists of 3 categorical IVs (1st IV consists of 2 categories, Second IV also consists of 2 Categories and 3rd IV consists of 3 categories).
How do we make a table for this including random effect parameters?
The *xtmixed* command has been renamed to *mixed* . You can fit a model with *mixed* and then use the *collect get* command to collect the results from the command. You would use a named expression with *collect get* to display the random effects as they are shown in the output from *mixed* . Then you would simply lay out your table with *collect layout* . Please email us at tech-support@stata.com if you would like to see an example.
This has simplified multilevel modelling
야호
Sir,
I am currently working on a meta analysis. But I am not able to produce association studies of allele contrast (A vs a). Except that all are done but can not understand how to calculate A vs a frequency. If you help me out it will help me a lot.
Thanks