Heteroscedasticity (Univariate models)

1. Normality, autocorrelation and heteroscedsaticity
Professor Sudarshan Bhattacharjee​ commented about diagnostic checking as such >> If
the data series is sufficiently large... about more than 30 0bservations then you can ignore
the violation of normality assumption coz by central limit theorem a long series will
converge to normality.... for auto-correlation and heteroskedasticity we can try first some
transformation in the variables.... however, if that does not work then use Newey-West
test for HC and HAC... This is readily available in R and Stata... Not sure about Eviews....
hope this helps..

2. What is heteroscedasticity?
Olasehinde Timilehin commented>Heteroscedasticity : remember that as your income
increases, the level of your hedonistic behavior tends to move along .As you are aging
over time, you will always remain the child of your parent whether they are rich or poor .
this is homoscedasticity....

3. Whether there is heteroscedasticity?
Arejib Ashraff posted the following digram.

Sheikh Muzammil Naseer commented> Hetroscedasticity is present...

Douma commented > the p value in less than 0.05, i think ther's heteroskedasticity.

Sayed Hossain commented> Null hypothesis: homoscedstic, Alternative hypothesis:
hetrocedasticity. As the p value is almost zero meaning we can reject null. So the
residuals are heteroscedastic.

3. It is the residual distribution of a regression liken taken from SPSS screen. Is
the model suffering from hetreroscedsaticity?
Sayed Hossain posted the figure below.

Muhammad Abid Economistc commented> I think Yes because of that point which is far
away from others known as outlier....

Muhammad Zhafran commented> Yes. There is a systematic pattern between the
residuals and predicted values. Am I right?

Khalid Mahmood Anjum commented> Yes systematic pattern exist in scatter plot so abid
and zhafran replied correctly

Sayed Hossain commented> Heteroscedasticity exists as there is systematic pattern or
specific direction movement in the residuals.

4. Please give me comment this output. I would be appreiciate any professors for
Khurshid Gelios‎ posted below.

Ghumro Niaz Hussain commented> do not reject null hypothesis at 5% and it is concluded
that variance of the residuals is homo....

Faridoon Khan Marwat Do not reject ur null hypothesis and turns to be homo

Sayed Hossain commented> We can not reject null meaning that residuals are
homoscedastic, which is desirable.

Muhammad Imran Javed commented> I agree with sir Syed Hossain and Khurshid Gelios
sir here you are doing hypothesis testing of hetero, not developing the relationship
between variables.
Like · Reply · November 26, 2016 at 2:55pm
Sayed Hossain

Saud Ahmad commnted> but we should not accept null if p-value is less than 10% level of
confidence. We can't ignore that. so I would suggest consider it rejection of null. as
usually we consider 1%, 5% and 10% level of significance for decision. Some economist
argue that in social sciences we should consider a var significant even at 12%.
Meo School of Research
Shishir Shakya
Noman Arshed
Hossain Academy Note
Univariate Models
Multivariate Models
Panel Data Model