ABSTRACT
Multiple Linear Regression (MLR) is a statistical procedure to estimate the functional relationship between a set of explanatory variables and a response variable. This is a cause and effect study. The partial influence of each variable on the response can be estimated from sample data. An example is to study “how socio economic factors and clinical parameters influence the quality of life in Chronic Kidney Disease (CKD) patients?”
Methods of fitting MLR model for a given data are discussed with computations in SPSS. Methods like stepwise, enter, remove, forward and backward for selecting independent variables into the model are discussed. The chapter also shows the method of testing the significance of the model by F-test, goodness of fit by R2 value and the importance of standardized regression coefficients. Model writing from the SPSS output and using it to predict the response/outcome value is stressed. Residual analysis and tests for normality of residuals using p-p plot are also highlighted and illustrated with live examples.(163)
