Unit Root Test in E Views| Panel Unit Root| Panel Data Analysis| E Views Tutorial | Stationarity

[@sabagulnaz592]

👍👍👍 If one variable say lngdp is stationery at level and another variable say lnpopulation is stationery at first difference, then can we proceed like this or do we have to do first difference of lngdp?

[@komalkanwarshekhawat_]

While reporting your results just mention that: lngdp is stationary at both level and at first difference. Rest of the variables have a unit root at a level and the absence of unit root at the first difference, indicating variables are integrated of I (1).

[@komalkanwarshekhawat_]

Also, before stationarity check, check cross-sectional dependence in your panels. If there is any cross-sectional dependence in your variables, apply second generation unit root tests Viz, CADF and CIPS.

[@sabagulnaz592]

Thank you😊

[@sabagulnaz592]

@@komalkanwarshekhawat_ how to do CADF and CIPS tests

[@komalkanwarshekhawat_]

@@sabagulnaz592 For that, first run Cross-sectional dependence test (Breusch pagan, pesaran CD etc) to find out whether there is any Cross-sectional dependence or not. If there is any, then you need to run second generation unit root test(CADF, CIPS) , this option is there only with unit root test. Thanks 😊

[@premsinghshekhawat3142]

Helpful 👍 Sharing it.

[@komalkanwarshekhawat_]

Thank you 🙏

[@amandeepkauruid1959]

mam … we may not say this every day but your inspirational words are like beautiful footprints that have been etched in our hearts and minds forever. Thank you

[@NATURE__Photography283]

👌👌👌❤

[@komalkanwarshekhawat_]

Thanks to you for all the Love and Support ❤️❤️ It's rare to find someone as generous as you. Keep learning 🌸

[@amandeepkauruid1959]

@@komalkanwarshekhawat_ its easy to find people like me but very difficult to find teacher like you mam

[@AsifAli-k6i8m]

best way of explaining statistical problems, keep it up.

[@komalkanwarshekhawat_]

Thank you so much 🙏😊

[@NATURE__Photography283]

Mam,you deserve all good and wealth in this world, because you love other as yourself and you always help and support. You have been a great teacher and lecturer to me, and I fully enjoyed your lectures, you explained everything in detail and it made it easier for me to understand.👍👍

[@amandeepkauruid1959]

😊😊🤗

[@komalkanwarshekhawat_]

Thanks a bunch for your kind words Sukhpreet ❤️❤️ You are the sweetest! Keep learning 🌸

[@simranpreetkaur236]

Mam you are perfect teacher 👍

[@komalkanwarshekhawat_]

No one is perfect ❤️ We strive to become better version of ourselves 🤗

[@aroraji5635]

Keep it up behan👌👌great work👍

[@komalkanwarshekhawat_]

Thanks bro ❤️

[@muhammadnaumansadiq]

Great Job👍👍👍

[@komalkanwarshekhawat_]

Thank you ☺️

[@bilqisbibi35]

Thank You Mam....

[@komalkanwarshekhawat_]

Your welcome .

[@meerulkataria6070]

Okay... Let me admit, this was just stupendous....❤️❤️🤗🤗... Aur agr stupendous se bhi aage hota hai toh woh hai apke efforts 😍❤️.... Be ready 🤪🤪🤪 bcoz m going to take economics optional for upsc...nd you are the one who have to guide me 🤪🤪🤪🤪🤪🤪😍❤️😍❤️

[@komalkanwarshekhawat_]

Thanks a ton ❤️ So, Economics is driving you crazy, I believe 😅. Keep going ❤️

[@bhartisingh5399]

Hardworking 🥰🥰🥰🥰🥰🥳

[@komalkanwarshekhawat_]

❤️❤️

[@kareenaa2656]

My inspiration ❤️

[@komalkanwarshekhawat_]

❤️❤️

[@dr.ejazanwer2493]

Doing well, Appreciable

[@komalkanwarshekhawat_]

Thank you for your kind appreciation!

[@deependrasinghrathore1785]

🔥🔥👍🏻👍🏻

[@komalkanwarshekhawat_]

Thank you 😊🤗

[@achrajkanwar3998]

Very informative 👍

[@komalkanwarshekhawat_]

Thank you 🙏

[@jeffarodi6300]

I am writing my Ph.D. thesis in social science, and your videos have completely made me abandon STATA. Thanks to you my data analysis job has been simplified. I have five variables and I opted to use only Levin-Lin-Chu's test, three of the variables are stationary at level, and the remaining two are at first difference. My question is? 1. Must I use the summary option to get results for all the test statistics available, despite having Levin-Lin-Chu's test in my proposal? 2. Considering that two of my independent variables only became stationary at the first difference, should I modify them when running the regressions?

[@komalkanwarshekhawat_]

Thanks for your kind appreciation. Regarding your queries- 1. Yes, it is always a good idea to support your stationary results with all the test statistics available. 2. It depends. If you are using a regression model/ Cointegration test wherein the pre requisite is that all the variables should be I(0) then you must ensure that. Otherwise, many regression models can be performed when the variables are I(0), I(1) or a mix of both. Good day!

[@omarmidi181]

You are great muah💋💋💋

[@vilchezalejandrianestoredu3]

gracias :)

[@komalkanwarshekhawat_]

Thank you. :)

[@bhattianamika]

Nicely done🥰

[@komalkanwarshekhawat_]

Thank you Anamika ❤️

[@AmandeepKaur-ij5zc]

Very nice mam 😇😇👍👍👍

[@komalkanwarshekhawat_]

Thanks dear 🤗🤗

[@iqbalhussain7268]

excellent way of teaching mam. if all the variables are non-stationary in the model, what we will do in the model estimation?

[@komalkanwarshekhawat_]

You can't employ regression model. It is pre requisite that all the variables should be stationary either at level or at first difference.

[@nourelhoudagorchene6628]

I reallyy appreciate your help in this subject. Could you please explain to me how I choose between the models ( model with intercept; model with intercept and trend or none )? which one is appropriate for this test please? and THANK YOUU

[@komalkanwarshekhawat_]

Thank you dear. If the results are not convincing in case of (model with intercept) you can report the results of (with trend). However, you can mention the results of both in your research. This only indicates that you checked the model with all the available options. Hope it helped. Keep following 🤗

[@nourelhoudagorchene6628]

@@komalkanwarshekhawat_ thaaank u so much, wish u the best ❤️❤️

[@ranjanachaudhari3206]

Hello maam Thank you for this great video so much helpful ! but i have one question as my one variable have unit root test even in the 2nd difference so what will be the next step do i test co integration and how to evaluate and check the cross sectional dependency? need your help

[@komalkanwarshekhawat_]

Hello Ranjana, mail your detailed query at [e-mail]. Since you have asked 3 things in the same question, this requires a detailed explanation. Good day 👍

[@renatasavira5128]

Hello Mrs. do you know how can i know f-stat and t-stat if i use FMOLS as my panel regression method? Because i have T>N so it force me to use non stationary panel data analysis beside the traditional one such as FEM,CEM OR REM

[@komalkanwarshekhawat_]

Dear, In case when T> N, one can use FMOLS, DOLS, Panel VECM etc. The t statistics will be obtained along with the regression coefficients result after employing the aforementioned models.

[@renatasavira5128]

@@komalkanwarshekhawat_ thankyou mrs for the answer. How about the f statistic if i use that model

[@komalkanwarshekhawat_]

@@renatasavira5128 The model is superior to OLS and provides with - t statistics, p values of regression coefficients, R square, adjusted R square and S.E of regression. Will upload a video soon on DOLS and FMOLS. Thanks.

[@VaishnaviBPHD-dn2lr]

Mam can you kindly tell how to add control variables in eviews

[@komalkanwarshekhawat_]

Please watch the video on GMM Model

[@mohdabdullah2088]

Please make video on how to arrange data and run GMM dynamic panel in eviews...

[@komalkanwarshekhawat_]

I have already uploaded these videos. Please check the playlist.

[@bellisma77]

Hi. Do we HAVE to test our series into the 3 options ( Individual intercept , Individual intercept and trend, or none) and then decide? Or if it was stationary in NONE option we conclude it is stationary? Thanx a lot

[@komalkanwarshekhawat_]

See, you can check for stationary (none, intercept, intercept and trend). Then, you can report the results as per the need. Some research papers report both- none, with intercept, some report only none results.

[@nabilahalhadisa7182]

Hi mam, thankyou for this great video. It's really helpful. But i have 1 question, i have independent variable, and 2 of them are dummy variable.when i try to run unit root test on my dummy variable, it's says "error unable to compute any results with selected option" can you help me to find the solution for my problem mam?

[@komalkanwarshekhawat_]

There is no need to run Unit Root Test for dummies.

[@eyuptanil4285]

are we apply SURADF second genetarion unit root test in stata?

[@komalkanwarshekhawat_]

Yes one can apply SURADF and SURKSS for second generation unit root test.

[@RinkuRinku-nl5zg]

👍👍👍

[@komalkanwarshekhawat_]

Thank you Shashi 😊🤗

[@muhammadzeeshan3735]

How to make data stationary from non stationary in Eview through command

[@komalkanwarshekhawat_]

Convert variables in log form.

[@khalidali2319]

👏🏻👏🏻

[@komalkanwarshekhawat_]

🤗😊

[@VaishnaviBPHD-dn2lr]

Mam how to check Wooldridge test for autocorrelation in panel data. please tell

[@komalkanwarshekhawat_]

Will soon upload a video on same. Keep following 😊

[@VaishnaviBPHD-dn2lr]

Thank you mam

[@afaq08]

I have eviews 10 version I can't found 2nd generation unit root tests

[@komalkanwarshekhawat_]

It is in EViews 12.

[@Mike_elGreco]

is it time series?

[@komalkanwarshekhawat_]

No, it is panel.

[@preetibedi5533]

what is the minimum number of years required to apply unit root for panel data in Eviews? my data is from 2015-2029 and 149 companies so it is 894 firm year observation but when I apply test it is saying insufficient number of observations.

[@komalkanwarshekhawat_]

The number of parameters should be less than number of cross sections.

[@preetibedi5533]

@@komalkanwarshekhawat_ what that means? Can you explain with simple example please That would really help And also number of years doesn't matter?

[@komalkanwarshekhawat_]

@@preetibedi5533 Increase time period or reduce number of companies

[@kinzawayne8376]

hi. my exogeneous variable is not becoming stationary in intercept; and intercept and trend at level; 1st difference; 2nd difference. it is only becoming stationary at none and 1st level difference. will this affect my analysis? please help

[@komalkanwarshekhawat_]

No it won't affect.

[@kinzawayne8376]

@@komalkanwarshekhawat_ thnx a lot

[@arshbunny6278]

💯🔥

[@komalkanwarshekhawat_]

😊🤗

[@aniksaha9925]

When i click unit root tests(from view), i find 2 options in eviews 12sv lite. A) cross sectionally independent, B) cross-sectionally dependent What to do

[@komalkanwarshekhawat_]

Watch the video on cross sectional dependence. If the variables are are cross sectionally dependent, then you need to run the second unit root test.

[@aniksaha9925]

Another query: what to do if a coefficient comes like this:" -1.41E-05"?

[@komalkanwarshekhawat_]

@@aniksaha9925 The result is not viable. There is High standard error may be.

[@aniksaha9925]

The corresponding p value is more than 5%, can i ignore this entity in writing regression equation?

[@maneesh123777]

I have panel data of 28 countries for 5 years. variables are diiferen in nature. 1st in percent, 2nd in rank (1-7), 3rd millions and 4th in billions. I have many questions. When I doing GMM i am getting result but I not able check causilty test. If I want to check long run and short run then which model I should use VAR,VECM, ARDL? As i know if unit root test of variables become stastionaly at different levels then we can't use VAR, VECM. In that case only ARDL Model can be used. I also want to how to determine which veriable in my case should be take in log form and which one not? Kindly help me. I can share my data if you want. Thanks

[@komalkanwarshekhawat_]

I would suggest few things - It is always great to set your panel data in the best way correctly. 1. 5 years duration is a very short time period to obtain robust results. 2. Please consider the variables in a common unit if possible. (Ex. If you are considering % , then take all possible variables in percentage) Moreover, you can watch videos on - GMM, VAR, VECM and ARDL . I have explained how and when to perform the test.

[@maneesh123777]

@@komalkanwarshekhawat_ I can understand time period is too short but its hard to find data on wb. my one variable is hightech import (% of total import) and availability of scientists and research in rank(1-7) as per wef standard. data for these 2 veriable not available more than 5 years.total patents are from few hunardes to millions. GDP in billion. I am trying to make 2 groups Asia and europe. then try to see the effect of htimport and Availability on patent and then effect og patent on gdp. Kindly guide me. If possible please provide your email.So i can send my data to you for review. Thanks

[@maneesh123777]

@@komalkanwarshekhawat_ in such case can i convert all variables into log form? one more thing i want to know. i got my gmm difference result. could you please suggest few other test to sustanciate my gmm result. such as lm etc. Thanks

[@komalkanwarshekhawat_]

@@maneesh123777 Sure it is [e-mail]

[@maneesh123777]

@@komalkanwarshekhawat_ Good Day! I have sent email to you. Once you get spare time then kindly reply. Looking forward to hearing from your side.Thanks, Maneesh

[@tosin_davidson]

Hello good day, Im writing a paper on Carbon emissions and I want to ask , waht do you do next if all the variables are stationary at first difference?

[@komalkanwarshekhawat_]

Run cross sectional dependence test

[@tosin_davidson]

@@komalkanwarshekhawat_ Oh okay, but do you have a video for it? Meanwhile, I was told to Pedroni cointegration test... Do u think that is also good?

[@komalkanwarshekhawat_]

@@tosin_davidson Yes. Video is already uploaded. Check playlist. Yes after cross sectional dependence test you can perform regression and Cointegration.

[@tosin_davidson]

@@komalkanwarshekhawat_ oh okay okay ... Thank you very much

[@nidhiverma3953]

👍👍

[@komalkanwarshekhawat_]

🙏😊