₦airaland Forum

Good day house. It's believed that when r squared is greater than durbin watson stat that the regression is spurious. How can one address a spurious reg result?

Battywonder:
Good day house. It's believed that when r squared is greater than durbin watson stat that the regression is spurious. How can one address a spurious reg result?

I think there is no how R-squared will greater than Durbin Waston.

R-squared value ranges from - 1 through 0 to (+) 1; while DW ran range from 0 to 4.

R-squared adjusted has be at least 0.5 which implies the independent variables explain 50% of factors that influence the model; while DW has to be in the neighborhood of 2 to show that the model does not suffer from serial correlation (i.e., error in past period doesn't correlate with another period)

To resolve spurious or nonsensical regression:

First, ensure that the time series or longitudinal data is stationary and the regression model was estimated at level or difference in which the data is stationary.

second, capture more independent variables that influence the explained variable.

third, increase your number of observation.

With those hints above, you should be able to resolve it.

Gracia

benebaby77:
I think there is no how R-squared will greater than Durbin Waston.

R-squared value ranges from - 1 through 0 to (+) 1; while DW ran range from 0 to 4.

R-squared adjusted has be at least 0.5 which implies the independent variables explain 50% of factors that influence the model; while DW has to be in the neighborhood of 2 to show that the model does not suffer from serial correlation (i.e., error in past period doesn't correlate with another period)

To resolve spurious or nonsensical regression:

First, ensure that the time series or longitudinal data is stationary and the regression model was estimated at level or difference in which the data is stationary.

second, capture more independent variables
third, increase your number of observation.

With those hints above, you should be able to resolve it.

Gracia

My model R-squared was 0.96 and DW 0.7 . Thanks for your recommendation as i am hopeful it will work.

Battywonder:
My model R-squared was 0.96 and DW 0.7 . Thanks for your recommendation as i am hopeful it will work.

Based on the two, your model suffer serial correlation and spurious estimate...


Try what I suggest and if it doesn't work.... mention me

Under cointegration test, if the trace statistics is greater than the critical value up to three cointegrating equations at 5% significant level, can we conclude that there is a unique long-run relationship among the variables under consideration?
benebaby77

Chylo:
Have a look at Microeconomic Theory by Mas-Colell and Intermediate Microeconomics by Hal Varian. These are the standard texts at the masters level too.

And to think I read it as an undergraduate...

Battywonder:
Under cointegration test, if the trace statistics is greater than the critical value up to three cointegrating equations at 5% significant level, can we conclude that there is a unique long-run relationship among the variables under consideration?
benebaby77

No, one has to compare the number of greater than and number of less than to determine if there is long run association in the variables in the model.

You can also consider running VECM to aid your analysis of cointegration....

benebaby77:
Based on the two, your model suffer serial correlation and spurious estimate...


Try what I suggest and if it doesn't work.... mention me

hello sir.. I need ur help on project topic selection..i just started 400l. We r required to use primary data..so i need something researchable.thanks.. Pls p.m me on kingnasss @ gmail.com or 0.8 1. 7.9 46 46.48..thanks...

kingnas:
hello sir.. I need ur help on project topic selection..i just started 400l. We r required to use primary data..so i need something researchable.thanks.. Pls p.m me on kingnasss @ gmail.com or 0.8 1. 7.9 46 46.48..thanks...

Every project topic is researchable. I would suggest you choose topic that seems to be your area of specializations/interests. There are many thousand topics you can select to research on.

I think if you are planning to specialize in a particular field of economics, I suggest you go for any topics that relate to that.

I will contact you soon via email.

Apart from Regression what else can Econometrics do?

25 most difficult questions to be prepared for on a job interview

Some ways have been proven to always win the heart of your interviewers

So, you are one of those who want to build a career and has several interviews to face in the coming months till you bag the best offer. Straightly, I will pour it out that the interview session is more than just questions and answers as we all know and the importance of equipping yourself before every interview no matter how much you know cannot be over emphasized.

RIDE WITH ME

http://jobintelng.com/2018/01/25-most-difficult-questions-to-be-prepared-for-on-a-job-interview/

binsidi:
Apart from Regression what else can Econometrics do?

Yes, although regression is the basic tool of econometrics, there are so.many types and techniques that attempts could be made to answer a multitude of possible economic questions.

Econometrics can be used to forecast or make predictions, develop economic models, analyse economic history and determine if and by how much something causes something else to happen. So, quite useful.

R square is to know how good fit a line is which ranges from 0 to 1 or 0% to 100%. R square is for regression and Durbin watson is to test for auto serial correlation which could be positive or negative depending of the range

Chylo:
Yes, although regression is the basic tool of econometrics, there are so.many types and techniques that attempts could be made to answer a multitude of possible economic questions.

Econometrics can be used to forecast or make predictions, develop economic models, analyse economic history and determine if and by how much something causes something else to happen. So, quite useful.

Yes it can forecast, models, analysis and causality as you mention above but all using regression still my question is not answered, and if you are thinking of techniques like tests e.g hetroskedasticity are all regression. Sir it cannot do anything apart from regression. If you did not agree please convinced me.

binsidi:
Yes it can forecast, models, analysis and causality as you mention above but all using regression still my question is not answered, and if you are thinking of techniques like tests e.g hetroskedasticity are all regression. Sir it cannot do anything apart from regression. If you did not agree please convinced me.

What about ANOVA analysis?

binsidi:
Yes it can forecast, models, analysis and causality as you mention above but all using regression still my question is not answered, and if you are thinking of techniques like tests e.g hetroskedasticity are all regression. Sir it cannot do anything apart from regression. If you did not agree please convinced me.

Basically, econometrics is about quantitative (and qualitative) modeling of economic theories and testing them with time series or panel data to see whether the empirical evidence obtained from the model conform with theories propounded. Results may vary with countries.

Majorly, in econometrics we make use of regression. But in contemporary econometrics, regression analysis has gone beyond the simple and multiple regression, there is VAR, VECM, cointegration, ARDL, GMM, GLM, probit, logit and tobit, ANOVA, Z-score, t-score etc....

There are many complex models used in financial econometrics but most of models have their foundation in regression...

Please I need you guys advice am a fresh economics graduate with a 2:1 and have the passion to establish a career in Investment banking/financial services currently planning to write a professional course during my nysc which of this Basic courses is more relevant as regard my career pursuit CIS/Ican

Micpopson:
Please I need you guys advice am a fresh economics graduate with a 2:1 and have the passion to establish a career in Investment banking/financial services currently planning to write a professional course during my nysc which of this Basic courses is more relevant as regard my career pursuit CIS/Ican

Go for ICAN sir and you wouldn’t regret it. ICAN gives you a wide range of knowledge across Finance- Financial Statement Analysis, Tax, Audit, Financial Management, Strategic Decision Making etc.

Wow, great thread! Great Economists!

Chylo:
What about ANOVA analysis?

Hi, are you done with ur phd

Hello guys, please keep up the good job.

Nevee:
Go for ICAN sir and you wouldn’t regret it. ICAN gives you a wide range of knowledge across Finance- Financial Statement Analysis, Tax, Audit, Financial Management, Strategic Decision Making etc.

If you are thinking of investment banking, go for CFA if you have sponsors but if u don't, just jeje go do CIS.

Hi chylo, I am a 2015 2:2 graduate of Economics based in abuja. Tired of job hunting with no luck. I want to enrol for an MSc in Political Economics and Dev. Studies Nasarawa State University, please any info if their form is still on sale and the prospects of this course as I initially wanted to go for an MSc in international econs, dev. Econs or research methodology

micflo28:
Hi chylo, I am a 2015 2:2 graduate of Economics based in abuja. Tired of job hunting with no luck. I want to enrol for an MSc in Political Economics and Dev. Studies Nasarawa State University, please any info if their form is still on sale and the prospects of this course as I initially wanted to go for an MSc in international econs, dev. Econs or research methodology

You can that find that out yourself by going through the website of your intending schools.

micflo28:
Hi chylo, I am a 2015 2:2 graduate of Economics based in abuja. Tired of job hunting with no luck. I want to enrol for an MSc in Political Economics and Dev. Studies Nasarawa State University, please any info if their form is still on sale and the prospects of this course as I initially wanted to go for an MSc in international econs, dev. Econs or research methodology

Hi micflo, sorry to hear about your job issues. Nigeria is not easy but we will overcome. I think either of any of the courses have prospects. Perhaps lecturing or research could be a direction, or international agencies and journalism. All the best and please keep us updated.

micflo28:
Hi chylo, I am a 2015 2:2 graduate of Economics based in abuja. Tired of job hunting with no luck. I want to enrol for an MSc in Political Economics and Dev. Studies Nasarawa State University, please any info if their form is still on sale and the prospects of this course as I initially wanted to go for an MSc in international econs, dev. Econs or research methodology

Please, can I have your contact for some enquiries. Thanks.

benebaby77:
Every project topic is researchable. I would suggest you choose topic that seems to be your area of specializations/interests. There are many thousand topics you can select to research on.

I think if you are planning to specialize in a particular field of economics, I suggest you go for any topics that relate to that.

I will contact you soon via email.

I pm you.

desoul2004:
I pm you.

Hi Desoul,

I don't have access to my mail again.....

What's good?

benebaby77:
Hi Desoul,

I don't have access to my mail again.....

What's good?

I just wanted to make some enquiries from you.