Descripción del título
Multivariate tests for time...
Review: The authors start out with stationarity testing, remarking that, as compared to unnivariate applications, multivariate applications are few. They fevelop the "joint stationarity" test [...]. In chapter 3, they turn to testing for "cointegration". When two nonstationary time series variables combine to generate a stationary error process in a regression, they are held to be "cointegrated". If such were found to be the case in the above example of GNP and Federal Spending, then the inference of a significant link between the two might not be spurios after all. The authors describe various cointegration tests, the most straightforward of which utilizes the well-known Durbin-Watson statistic.[...] The final chapter gives helpful guidance on computer software. [Fuente: Michael S. Lewis-Beck]
Monografía
monografia Rebiun35318065 https://catalogo.rebiun.org/rebiun/record/Rebiun35318065 190318s1995 usad fr 001 0 eng d 0-8039-5440-9 UAN0103246 UAN eng 519.237 303.7.032.4 519.246.8 001.89::303 Multivariate tests for time series models Texto impreso] Jeff B. Cromwell, Michael J. Hannan, Walter C. Labys, Michel Terraza. Thousand Oaks (California) London New Delhi Sage cop. 1995. Thousand Oaks (California) London New Delhi Thousand Oaks (California) London New Delhi Sage VI, 98 p. gráf. bl. y n. 22 cm VI, 98 p. Sage university papers. Quantitative applications in the social sciences 07-100 Bibliografía: p. 91-96. [Chap.] 1-8: Introduction ; Testing for joint stationarity, normality, and independence ; Testing for cointegration ; Testing for causality ; Multivariate linear model specification ; Multivariate nonlinear models ; Model order and forecast accuracy ; Computational methods for performing the tests -- Appendix: Statistical tables Review: The authors start out with stationarity testing, remarking that, as compared to unnivariate applications, multivariate applications are few. They fevelop the "joint stationarity" test [...]. In chapter 3, they turn to testing for "cointegration". When two nonstationary time series variables combine to generate a stationary error process in a regression, they are held to be "cointegrated". If such were found to be the case in the above example of GNP and Federal Spending, then the inference of a significant link between the two might not be spurios after all. The authors describe various cointegration tests, the most straightforward of which utilizes the well-known Durbin-Watson statistic.[...] The final chapter gives helpful guidance on computer software. [Fuente: Michael S. Lewis-Beck] Donación Arthur J. Kendall. Análisis multivariante Series temporales Ciencias sociales- Investigación Ciencias sociales- Métodos estadísticos Cromwell, Jeff B. Hannan, Michael Joseph. Labys, Walter C. 1937-) Terraza, Michel.