Jarque Bera PDF
Jarque Bera PDF
Jarque Bera PDF
We use the Lagrange multiplier procedure to derive efficient joint tests for residual
normality, homoscedasticity and serial independence. The tests are simple to compute and
asymptotically distributed as x2.
1. Introduction
01651765/81/0000-0000/$02.50 0 North-Holland
256 C.M. Jurque, A. K. Benz / Tests for normali~v, homoscedustxity, serial independence
(1)
4c,,c2,u2,~,~) = -
*=,-00 1
l$ln[ Jm ewbk)% + i
t=l
+k>. (2)
’ Our test is derived for the Pearson Family and hence, for this, we know it will have
optimal properties. However, the test is not restricted to members of this family.
C. M. Jarque, A.K. Beru / Tests for normality, homoscedustiaty, serial independence 251
* The proof of this and related results is available from the authors
258 C. M. Jarque, A. K. Bera / Tests for normality, homoscedasticity, serial independence
first term in (4) is identical to the LM residual normality test for the case
of HI residuals [e.g., Jarque and Bera (1980)], say LM,. The second term
is the LM homoscedasticity test for the case NI residuals [e.g., Breusch
and Pagan (1979)], say LM,. Finally, the third term is the LM serial
correlation test for the case of NH residuals [e.g., Breusch (1978)], say
LM,. Therefore, it is interesting to note that these ‘one-directional tests’
may be combined as above, to obtain a joint test for residual NHI.
LMNHI is, under H,, asymptotically distributed as x;+~+~ and H,
would be refitted if the computed value of LM,, exceeded the desired
significance point on the x:+~+~ distribution. The test is asymptotically
equivalent to the likelihood ratio test and this implies that it has, for
large-samples, maximum local power. In addition, it is simple to compute
requiring only OLS residuals. As a result, it should prove to be a useful
diagnostic check in the analysis of regression residuals.
3. Concluding remarks
References
Durbin, James and G.S. Watson, 1950, Testing for serial correlation in least squares
regression I, Biomettika 37, 409-428.
Jarque, Carlos M. and Anil K. Bera, 1980, An efficient large sample test for normality of
observations and regression residuals, Paper submitted to the Journal of the American
Statistical Association.
Kendall, Maurice G. and Alan Stuart, 1969, The advanced theory of statistics, Vol. 1.
(Griffin, London).
White, Halbert and Glenn M. MacDonald, 1980, Some large sample tests for non-normality
in linear regression models, Journal of the American Statistical Association 75, 16-28.