ABSTRACT
This chapter describes the methods for dealing with multivariate data. A variable is continuous if it is measured on a scale like length, time or weight, where any values between points on the scale are possible. A variable is discrete if only certain values are possible, for instance a count of interactions can take only whole number values and so is discrete. A population denotes all members of some category of interest, which may be people, other entities or a reference set of potential observations. Multiple regression extends the method to include several IVs. Generalized Linear Models extend further the idea of the General Linear Model. Factor analysis is used a lot in psychology, either in an exploratory way, as in attempts to identify an underlying factor that cannot be measured directly but influences scores on several variables, or in a confirmatory manner, as in testing a model that specifies particular relationships among latent variables and their measurable indicator variables.
