Factor analysis is the generic term given to a group of multivariate statistical techniques that can be used to identify, from a large pool of variables, the important variables related to a study. Variables that are interrelated in terms of their correlations can be grouped, thereby possibly identifying some otherwise hidden concept(s). The variables that group or cluster together can be combined into a weighted linear combination called a factor, which can then be used to measure concepts not otherwise directly observable. Thus factor analysis can facilitate the reduction of focus to a smaller set of variables, facilitate concept identification and facilitate measurement of these concepts.

The various techniques of factor analysis differ in the manner in which the weights for the variables constituting the factors are determined. The most widely used method for determining factors is that of principal component analysis, which is the method used in analysing the responses to the user satisfaction questionnaire as described in Chapter 13. In the sequel the term factor analysis is used as a synonym for principal component analysis.

Factor analysis can be used both for exploratory and confirmatory studies. The use of factor analysis for concept identification illustrates how the technique can be used in an exploratory manner. It is used in a confirmatory sense when, for example, the factor analysis solution is compared to some a priori postulation concerning the grouping of the variables into factors. In our study it was used for exploratory purposes.

Once the analyst...