Testing the normality of residuals on regression model for the growth of sludge microbes on PEG 600

Abstract

Polyethylene glycols (PEGs) are employed in numerous sectors. PEGs are nephrotoxic and their biodegradation by microbes could be a potential tool for bioremediation. Numerous bacterial growth studies neglect primary modelling even though modelling exercises can reveal important parameters. Previously, we have utilized several growth models to model the growth of sludge microbes on PEG 600. We discovered that the modified Gompertz model via nonlinear regression utilizing the least square method was the best model to describe the growth curve. However, the use of statistical tests to choose the best model relies heavily on the residuals of the curve to be statistically robust. Normality tests for the residuals used in this work has indicated that the use of the modified Gompertz model in fitting of the growth curve of the sludge microbes on PEG 600 initially was  not adequate due to the presence of an outlier. Upon removal of this outlier, the residuals conformed to normality test, visually and statistically