Measurement of Action Result of Eliminating Bivariate based Cook’s D: Cell Concentration and Turbidity

Abstract

The method of turbidimetry interprets the effects of light scattering of the cell particles consisting of the mean of cell concentration. The relation between number of cell particles and turbidity has not been adequately expressed particularly to bivariate statistically. The objective of this paper is to discuss the action result of eliminating a bivariate-based Cook’s D analysis, which can be measured with relative error of cell quantification
in turbidimetry. Experiments were designed to research the regression design on two paired variables (bivariate), and according to the linear relationship characteristic between suspended solid and turbidity. Variables consisted of cell concentration and turbidity. Paired sample size was ten data as a dataset and replicated five times. The diagnosis of regression on linearity parameter against five sets of data was used as assumption of covariance analysis, backward method–residual analysis of Cook’s D, and then measured by relative error value of cell concentration result by a turbidity value. The regression diagnostic of plotting fifty bivariate had the heterochedastisity character with five slopes being not homogenous. The diagnosis a linier curve was performed and elimination action toward to a bivariate at the furthest point using Cook’s Distance value. One
dataset gives the lowest relative error result of cell quantification and become consideration in turbidimetry as a standard curve.