Panel Outlier Robust / Quantile Regression with Fixed Effects for Static Panel Data in STATA
HTML-код
- Опубликовано: 19 окт 2024
- When the #paneldata is not normally distributed mainly because of presence of #outliers, we have to resort to #median regression. This tutorial describes the panel quantile regression #PQR which is instrumental in #non-normal variable or variables with outliers cases. This method can also generate coefficients at different specified #percentile position to conduct a #dataanalysis which assess differences in effect at different #distribution position.
Informative. Well done brother
Thanks
Thank you for the video.
Hi Sir,
I have 456 obs of 24 countries over 19 years. I get different coefficients and p values everytime I run qregpd for all quantiles.
Could you please help how to resolve the issue?
Rqpd uses the random number to start search for coefficients. Since the data distribution may be nonlinear so differenr random number may provide different optimal solution. So you should write set seed 123 before the rqpd command so that it fixes the random number generator algorithm to provide you same results every time you estimate the command again. Here the number 123 is just a random number you can use any
Good day sir, I’m facing a similar challenge here. Please what do you mean by “write set seed 123” I don’t understand. I have only a month left to present my thesis, please help me.
@@isaacbekoe657 It is a random number generator command, when you write set seed 123, it uses 123 as a start of random number algorithm which is used to estimate the PQR regression. If you do not use the set seed command before PQR it will change the estimation results every time it is estimated
thank you
Welcome!
Thanks for your supports, Sir. Please do video on GMM Quantile regression.
As soon as possible
Hello! What if I need to use bootstrap in the Panel Quantile Regression? How should I proceed?
It the manual the term is bsqreg which could be used for non panel data further the panel quantile regression codes could also include bootstrap.
It the manual the term is bsqreg which could be used for non panel data further the panel quantile regression codes could also include bootstrap.
@@nomanarshed I see. Thanks for your reply
Thank you. Where can I get this datafile?
Please is it wrong for me to use the sqreg command to run quantile regression panel data model? Moreso is like i get all my variables to be significant when I use your command, I’m a bit skeptical about that. Any help?
Study the base paper which has introduced the sqreg to find guidance
Hİ, sir. Do we test other hypothesis for example normality, heteroskedasity? And I sent a e mail. Please I need for help so may you back for me? Thanks.
For normality sktest and other commands can be used. For heteroskedasticity there are tests developed for each model, or you can manually generate residuals and check heteroskedasticity using BGP method.
hello, how do we know when to use arate 0.5 and arate 1 when running qregpd?
Usually start with 0.5 and then make plots for all cases.
well done work,
bro could you please make a video how to plot Q.regression coefficients
You can learn here - ruclips.net/video/RikGNzujJr4/видео.html
From your video I didn't see the Constant Term. Is panel quantile regression able to run the constant term? also where could I find the Goodness of Fit in this? ....is it the Mean acceptance rate? thanks in advance
Since its a fixed effect model the intercept is varying across cross sections. It can be visualized for that the documentation of that command must be studied. Further as per my knowledge we need to manually generate the goodness of fit by comparing Y and Yhat.
@@nomanarshed thank you for your advice.
Hello sir i am doing analysis with the help of qregpd but i want to draw a graph please share a command for qregpd plot in your previous video you discuss it but its to complicated please help me
These graphs are intrepreted using eviews in link.springer.com/article/10.1007/s11205-022-02906-9
Sir you share a link for qregpd plot but I don't understand sir please share a simple command or method for qregpd plot thank you
Hello sir I am doing analysis already but qregpd plot are remaining you share with me a link of video but its too complicated so send me a simple method and command thanks
@@shayankhan1264 you can extract the codes using ruclips.net/video/7tCTaO2hGgY/видео.html
Dear Sir,
How do we plot the tables. I mean how we derive graphs from the tables? Thanks.
Syed Sadaqat, PhD
Within stata there is not builtin command like eviews. Either you have to store the coefficients quantilewise in a new variable using hardcore coding and then draw graphs. Or just type them in excel and draw simple graphs
Can you please share the data for trying Quantile Regression Lecture cmds once again please
You can replicate codes using this ruclips.net/video/7tCTaO2hGgY/видео.html while running the video in high definition
Please do video on how to plot graph of panel quantile regression in Stata.
It is shown here ruclips.net/video/RikGNzujJr4/видео.html
This command is based on which paper?. And how it incorporates country fixed effect.
These commands are based on self research of STATA documentation. The paper reference can be found in the documentation. This tutorial is fixed effect based but can be altered by excessing help
@@nomanarshed can we apply panel Quantile regression when t=7 and N=80?
There are various papers on quantile panel regression. Eg. Powell (2016). So , i just want to know that this command will estimate which model
command sec is not recognized is showing after i enter the command
You can increase the quality of the video to read the commands
Hello, Thanks for your supports.
At the end of the video I added the phrase (quantile(25)), what does it mean? How do you determine the number (25), and what does it mean?
25 means that it will select 25th percenile value as center to determine point estimate
@@nomanarshed
25
Is it Median value?
How do you determine it?
@@salahnasr5195 it is 25th percentile, i have not determined it, it is the commend to show that outcome, you can change it to 50 if you want to see results at median
@@nomanarshed
Thank you for this offer. If you have a paper for the same methodology please attach it.
@@salahnasr5195 link.springer.com/article/10.1007/s11205-022-02906-9
can we see the graph
Which graph