Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

Violations of assumptions of

the CLRM
Violation of the assumption of normality
of errors
• One of the assumptions of the CLRM is that the
disturbance terms are normally distributed.

• The implication of the violation of this assumption is that


the estimators of the regression coefficients are also not
normally distributed.

• Consequently, the F-test of model adequacy and the t-


test of significance of regression coefficients will not be
applicable.
• The Jarque-Bera test is one of the common tests of
normality.

• This test is based on the coefficients of skewness (S)


and kurtosis (K) of the residuals.
• The other characteristic of a distribution is its
peakedness which is measured through the coefficient
of kurtosis.

• A distribution is said to be leptokurtic if it exhibits fat


tails and excess peakedness at the mean - the
coefficient of kurtosis by far exceeds three.

• A distribution which has a smaller peak at the center is


called platykurtic - the coefficient of kurtosis is smaller
than three.
Test of normality of errors in EViews
To obtain the fitted regression model, click on:

Quick Estimate Equation…


In the Equation Estimation window, type:
imports c gdp stock_formation
When you click on OK, you will get the fitted multiple linear
regression model with several model-fit statistics.
To view the results of the test of normality of residuals, in the
equation window click on:

View Residual Diagnostics Histogram-Normality Test


You will automatically get the results of the Jarque-Bera
test of normality
• The Jarque-Bera test statistic is 1.119511 with a p-
value of 0.571349.

• Since the p-value exceeds 0.05, we do not reject the


null hypothesis of normality of errors.

The rule of thumb seems to be:


1) If the skewness is between -0.5 and 0.5, the data are
fairly symmetrical.
2) If the skewness is between -1 and – 0.5 or between
0.5 and 1, the data are moderately skewed.
3) If the skewness is less than -1 or greater than 1, the
data are highly skewed.

You might also like