The analysis of multivariate models is of great importance within econometrics and other social sciences. Allocation models (e.g., demand and production functions), panel data models, and simultaneous demand and supply models are all examples of such models that involve systems of equations. The building of reliable econometric models in these situations is thus a central area of research.

The assumptions that lie behind such econometric models must however be valid, at least approximately, if the model is to be of any practical use. Testing these assumptions (so called misspecification testing) is thus one of the central pillars of econometric model building, a point that has been emphasised more and more during the last twenty years, especially by the LSE school. The systematic application of misspecification testing has only sporadically been applied to multivariate models, however.


Papers from the Statistical Research Unit/Dept. of Statistics, Univeristy of Gothenburg

Frisén, M. (2007) Ed. Financial Surveillance. Wiley.

Lindström, F. (2003): On prediction accuracy of the first order vector auto-regressive process. Research Report 2003:7, Statistical Research Unit, Göteborg University. Licentiate thesis.

Holgersson, T. (2003): On Assesing Distributional Properties of Multidimensional Variables. Publications no 28, Statistical Resarch Unit, Göteborg University. Dissertation thesis.

Shukur, G. & Edgerton, D. (2002): The Small Sample Properties of the RESET Test as Applied to Systems of Equations. J. of Statist. Comput. Simul. 72(2): 909-924.

Holgersson, T. (2001): On assesing multivariate normality. Research Report 2001:1, Department of Statistics, Göteborg University. Licentiate thesis.

Holgersson, T. (2003): Testing for multivariate autocorrelation. Research Report 2003:2, Statistical Research Unit, Göteborg University.

Holgersson, T. & Shukur, G. (2003): Testing for multivariate heteroscedasticity. Research Report 2003:1, Statistical Research Unit, Göteborg University.

Holgersson, T. (2002): Testing for non-normality in multivariate regression with nonspherical disturbances. Research Report 2002:9, Department of Statistics, Göteborg University.

Holgersson, T. and Shukur, G. (1999): An investigation of the properties of some frequently used normality tests in a system perspective. Department of Statistics, Göteborg University, Sweden. (Working in progress.)

Almasri, A. and Shukur, G. (1999): Using wavelet analysis and bootstrap technique in investigating the causality relations in integrated-cointegrated VAR systems. Department of statistics, Göteborg University, Sweden. (Working in progress.)

Edgerton, D. and Shukur, G. (1999): Testing autocorrelation in a system perspective, Econometric Reviews, 4.

Manatalos, P. and Shukur, G. (1999): Testing for cointegrating relations: A bootstrap approach. Research Report 1999:5, Department of Statistics, Göteborg University, Sweden. Submitted for publication, Communications in Statistics.

Shukur, G. (1999): The effect of non-normal error terms on the properties of the systemwise RESET test. Research Report 1999:6, Department of Statistics, Göteborg University, Sweden. Submitted for publication, Computational Statistics & Data Analysis.

Shukur, G. (1999): Att utforska ekonometriska modellers förmåga att beskriva verklighet. Qvartilen, Svenska statistikersamfundets medlemsblad. Årgång 14, nr. 1, mars 1999. A note.

Manatalos, P. and Shukur, G. (1998): Size and power of the error correction model (ECM) of cointegration tests. A bootstrap approach. Oxford Bulletin of Economics and Statistics, 60.

Hatemi-J, A. and Shukur, G. (1999): The causal nexus of government spending and revenue in Finland: A boostrap approach. Applied Economics Letters, 6.

Shukur, G. (1998): The robustness of the systemwise Breauch-Godfrey autocorrelation test for non-normal error terms. Research Report 1998:11, Department of Statistics, Göteborg University, Sweden. Accepted, under revision for publication, Communications in Statistics.