00000nam 2200000za 4500 9.847756 CaOODSP 20241203113640 cr ||||||||||| 171124s1984 onca|||fo f000 0 eng d CaOODSP eng eng fre n-cn--- En13-5/85-21E-PDF Halfon, Efraim. The bootstrap and quantitative structure activity relation multiple linear regression models for nonparametric estimates of standard error [electronic resource] / by Efraim Halfon. Burlington, Ont. : Aquatic Physics and Systems Division, National Water Research Institute, Canada Centre for Inland Waters, 1984. 8, [7] p. : ill. NWRI contribution ; 85-21 Digitized edition from print [produced by Environment and Climate Change Canada]. "November 1984". Includes bibliographical references. "A recently invented statistical method, the bootstrap, isused to verify whether a multiple regression model, developed from a limited data set, produces reliable predictions of toxicity if all possible data would have been available to develop the model. This statistical method allows generalization to chemicals of the same class not included in the original analysis. An example is taken from the literature and the errors associated with the coefficients of a multiple regression are computed using the bootstrap. Also, the multiple regression model is used to estimate the toxicity of a contaminant not used in model development and estimates with their relative probability are graphically displayed"--Abstract. Includes abstract and executive summary in French. gccst Water quality gccst Statistical analysis Canada. Environment Canada. National Water Research Institute (Canada) NWRI contribution ; no. 85-21 (CaOODSP)9.844121 PDF 1.05 MB https://publications.gc.ca/collections/collection_2017/eccc/En13-5-85-21-eng.pdf