Testing the effect of blocking in a randomized complete block design (RCBD)

Autor:	Clovis A. Peres
Rok vydání:	1981
Předmět:	Statistics and Probability F-test Generalized randomized block design Statistics Randomized block design Covariance Constant (mathematics) Blocking (statistics) Approximate solution Mathematics Block (data storage)
Zdroj:	Communications in Statistics - Theory and Methods. 10:2447-2459
ISSN:	1532-415X 0361-0926
DOI:	10.1080/03610928108828201
Popis:	In the RCBD, the block system causes, for most applications, strong correlation between responses from experimental units in the same block. Therefore, the usual model with the usual assumptions about zi j. (i.i.d. N(0,σ2)) is not realistic. In this paper we study the consequences of more realistic assumptions about zi j. and verify that the usual F test, for testing treatment effects, can be used even if there is a constant covariance between responses from experimental units in the same block. When the variances associated with each treatment are not the same we give an approximate solution for testing treatment effects.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::1124035df25987dc6a5ff62e708d6cdf https://doi.org/10.1080/03610928108828201 Zobrazit plný text záznamu