ABSTRACT
This chapter describes the rationale behind factor analysis and presents an example of its use. It provides a general understanding of the mechanics of factor analysis and its interpretation. Factor analysis is a statistical means of reducing the redundancy in a large data set by organizing the data in terms of shared variability among the measures. The systematic differences among the products on a given sensory attribute or chemical compound constitute the variance of that attribute or compound. Overlap of the circles represents a correlation or covariance between the variables, with the degree of overlap representing the degree of correlation. The value of factor analysis and principle component analysis lies in its ability to parsimoniously summarize data sets which consist of many variables and are therefore unwieldy. The correlation matrix is the starting point input to a factor analysis. There are numerous methods available for factoring a correlation matrix.
