Trend analysis using a multivariate procedure for data with unequal residual covariance and regression coefficient matrices: Application to Canadian Atlantic cod contaminant data

Autor:	Charles J. Musial, R. K. Misra, C. L. Chou, D. P. Scott, John F. Uthe
Rok vydání:	1993
Předmět:	Analysis of covariance Multivariate statistics Multivariate analysis Statistics Linear regression Univariate Aquatic Science Covariance Oceanography Residual Pollution Regression Mathematics
Zdroj:	Marine Pollution Bulletin. 26:73-77
ISSN:	0025-326X
Popis:	Conventional univariate and multivariate analysis of covariance (ANCOVA and MANCOVA) are commonly used to investigate time trends in contaminant levels in fish tissues. Both procedures require that years do not differ in their residual variances, covariances, and regression parameters. These requirements were not met in many data sets. A multivariate procedure that does not have these requirements was used to reanalyse contaminant concentration trends in Canadian cod data collected between 1977 and 1985.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::37abbabc5f21c663d8d4ff9b1e6de947 https://doi.org/10.1016/0025-326x(93)90094-z Zobrazit plný text záznamu