ABSTRACT
Analysis of Variance (ANOVA) is familiar tool for comparing the mean responses of three or more groups of cases found in a survey/experiment. The unified approach of ANOVA and Linear Regression allows one to incorporate one or more continuous covariates and estimate the response after adjustment for covariates.
Starting with a brief review of univariate ANOVA this chapter explains methods and skills of handling more than one factor and their interactions on the response. The aim is to alert the researcher on testing how the mean response changes with a group of factors (having few discrete levels) and their interactions on the outcome of the study and test for their statistical significance. The use of general linear model and its R2 value in estimating the response is highlighted and demonstrated in SPSS environment with real data. Special focus is given on handling continuous covariates leading to Analysis of Covariance and the use of adjusted means is also highlighted. Illustrations show how to interpret the output of SPSS. (167)
