Investigating Multicollinearity in the Environment and Health Study

Human biomonitoring is the measurement of chemicals in blood, urine, and other tissues or fluids. It is becoming a common tool in the study of human exposure to environmental pollution and the potential health effects of those pollutions. This work reveals the correlation between the measured markers of internal exposure and techniques to deal with them. A number of diagnostics tests used to determine the degree to which multicollinearity might exist includes correlation matrix analysis, regression analysis, investigating the stability of the regression model, variance inflation factors (VIFs), the tolerance and condition indices. Factor analysis and ridge regression were considered as remedial measures for multicollinearity. This book will be useful to Biostatisticians, Health and Environmental workers or researchers who might want to study the inimical effects of environmental pollution to human health.