Practitioners corner a positive semi-definite covariance matrix for Hausman specification tests of conditional and marginal densities

Article Abstract:

A semi-definite covariance matrix for the use of Hausman specification tests to identify misspecifications of the conditional and marginal densities in finite samples is introduced. The matrix address the problem arising from a negative test statistic. The latter occurs when the estimated covariance of the difference of the two maximum likelihood estimators in the Hausman test are negative semi-definite in finite samples.

Author: Ai, Chunrong
Research, Density functionals, Density functional theory, Maximum likelihood estimates (Statistics), Analysis of covariance, Covariance analysis, Maximum likelihood (Statistics)

User Contributions:

Comment about this article or add new information about this topic:

CAPTCHA


On the power of GLS-type unit root tests

Article Abstract:

A variation of the Dickey-Fuller unit root test is presented. Non-stochastic components are generally not more accurate under this estimate.

Author: Taylor, A.M. Robert, Burridge, Peter
Statistical Data Included, Models, Testing, Stochastic analysis, Mathematical models, Linear models (Statistics)

User Contributions:

Comment about this article or add new information about this topic:

CAPTCHA


Panel unit-root tests for cross-sectionally correlated panels: a Monte Carlo comparison

Article Abstract:

Performance of Monte Carlo comparison for comparison of panel unit-root tests for cross-sectionally correlated panels, is discussed.

Author: Gutierrez, Luciano
Italy, Data Processing, Methods, Usage, Statistical methods, Electronic data processing

User Contributions:

Comment about this article or add new information about this topic:

CAPTCHA

This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.