ABSTRACT
Factor analysis begins with a data matrix. Each variable in the data matrix has a particular mean, standard deviation, and distribution. Variations in the characteristics of the variables may have a definite effect on a factor analysis. One instance in which the distributions may distort the factor analysis is if the variables are dichotomous rather than approximations to continuous variables. The problems and possibilities of factor analysis with noncontinuous variables are discussed in Section 14.1.
