Biometrical Letters vol. 43(2), 2006, pp. 99-107
|
COMPARISON OF SEVERAL TEST PREPARATIONS WITH ONE STANDARD IN MULTIVARIATE BIOASSAYS Zofia Hanusz1, Agnieszka Rutkowska2 1Department of Applied Mathematics, Agricultural University, Akademicka 13, 20-950 Lublin, Poland, zofia.hanusz@.ar.lublin.pl 2Institute of Soil Science and Plant Cultivation, National Research Institute in Pulawy, Poland, agrut@iung.pulawy.pl |
In the paper we consider the problem of estimation of the relative efficiencies of several test preparations with respect to one standard preparation. We consider the case where several doses of each preparation are administered to units forming block designs. We assume that the effect of doses of preparations is measured by several measurable traits forming multivariate responses. We assume that responses can be described by a parallel-line model with independently and normally distributed vectors of errors with the same covariance matrix. The statistical method presented in the paper is applied to a comparison of the impact of different forms of nitrogen fertilizer on yield of winter wheat.
relative potency, multivariate bioassays, parallel-line assay, multivariate observations, block designs, testing hypotheses, estimation of parameters.
