White''s general heteroscedasticity test, Advanced Statistics

The Null Hypothesis - H0:  γ1 = γ2 = ...  =  0  i.e.  there is no heteroscedasticity in the model

The Alternative Hypothesis - H1:  at least one of the γi's are not equal to zero i.e. the squared residuals are related to one of the independent variables.

Reject H0 if nR2   >   2094_Tests for Heteroscedasticity.png

Regression Analysis: sqresi versus totexp, age, ...

* sqnk is highly correlated with other X variables

* sqnk has been removed from the equation.

 

The regression equation is

sqresi = 0.00086 - 0.000117 totexp + 0.000765 age + 0.00007 nk

         + 0.000000 sqtotexp - 0.000009 sqage + 0.000000 totexpage

         + 0.000026 totexpnk - 0.000077 agenk

 

Predictor         Coef     SE Coef      T      P     VIF

Constant      0.000857    0.007288   0.12  0.906

totexp     -0.00011676  0.00004906  -2.38  0.017  49.148

age          0.0007649   0.0003256   2.35  0.019  73.466

nk            0.000072    0.002941   0.02  0.980  24.250

sqtotexp    0.00000019  0.00000010   2.00  0.045  13.958

sqage      -0.00000879  0.00000394  -2.23  0.026  62.515

totexpage   0.00000021  0.00000097   0.21  0.831  37.830

totexpnk    0.00002598  0.00001464   1.77  0.076  18.920

agenk      -0.00007694  0.00007905  -0.97  0.331  32.566

S = 0.0112807   R-Sq = 1.2%   R-Sq(adj) = 0.6%

 

Analysis of Variance

Source            DF         SS         MS     F      P

Regression         8  0.0022313  0.0002789  2.19  0.026

Residual Error  1493  0.1899897  0.0001273

  Lack of Fit    639  0.0804237  0.0001259  0.98  0.601

  Pure Error     854  0.1095659  0.0001283

Total           1501  0.1922209

 

 332 rows with no replicates

Source     DF     Seq SS

totexp      1  0.0006642

age         1  0.0000000

nk          1  0.0000026

sqtotexp    1  0.0005240

sqage       1  0.0005895

totexpage   1  0.0000013

totexpnk    1  0.0003292

agenk       1  0.0001206

 

MTB > let k4=1502*0.012

MTB > print k4

Data Display

K4    18.0240

Inverse Cumulative Distribution Function

Chi-Square with 8 DF

P( X <= x )        x

       0.95  15.5073

Since nrsq = (1502*0.012) 18.024 > 15.5073 = 2094_Tests for Heteroscedasticity.png, there is sufficient evidence to reject H0 which suggests that there is heteroscedasticity in the model from White's general heteroscedasticity test at the 5% significance level.  Both Breusch Pagan test and White's general heteroscedasticity test seem to indicate that totexp is the culprit as the T value is significant and the P-value is 0.000.

Posted Date: 3/5/2013 6:04:19 AM | Location : United States







Related Discussions:- White''s general heteroscedasticity test, Assignment Help, Ask Question on White''s general heteroscedasticity test, Get Answer, Expert's Help, White''s general heteroscedasticity test Discussions

Write discussion on White''s general heteroscedasticity test
Your posts are moderated
Related Questions
Different approaches to the study of early indian history

You may have the opportunity to buy some electronic components. These components may be reliable (1) or unreliable (2). The potential pro?ts are £10,000 if the components are rel

Paired availability design  is a design which can lessen selection bias in the situations where it is not possible to use random allocation of the subjects to treatments. The desig

Graphical deception : Statistical graphics which are not as honest as they should be. It is relatively simple. To mislead the unwary with the graphical material. For instance, c

Mendelian randomization is the term applied to the random assortment of alleles at the time of gamete formation, a process which results in the population distributions of genetic

Canonical correlation analysis : A process of analysis for investigating the relationship between the two groups of variables, by ?nding the linear functions of one of the sets of

Centile reference charts : Charts which are used inmedicine to observe the clinical measurements on individual patients in the context of the population values. If the population i

Missing values : The observations missing from the set of data for some of the reason. In longitudinal studies, for instance, they might occur because subjects drop out of the stud

The Null Hypothesis - H0:  There is no heteroscedasticity i.e. β 1 = 0 The Alternative Hypothesis - H1:  There is heteroscedasticity i.e. β 1 0 Reject H0 if Q = ESS/2  >

The variables appearing on the right-hand side of equations defining, for instance, multiple regressions or the logistic regression, and which seek to predict or 'explain' response