ABSTRACT
In this and the next two chapters I shall take some examples of published factor analyses and subject them to detailed scrutiny. I shall highlight those aspects of the published papers which are essential to the proper use of factor analysis. These are as follows.
The questions which the research was trying to answer. This is a vital aspect, as should be obvious of any research. Yet all too often, factor analyses are conducted almost for their own sake: to many researchers factor analysis is the Everest of techniques which must be climbed because it is there in the computer package.
In this and the next two chapters I shall attempt to indicate what problems factor analysis is suited to solve and conversely what it is unable to deal with. Powerful as the technique is there are questions which are more efficiently answered by other methods.
The factor analytic methodology. Here I shall discuss the technical aspects of the factor analysis, most of which have been explicated in previous chapters. I shall carefully go through the sampling of subjects and variables and the type of factor analysis and rotation to see to what extent the results are likely to be distorted by the methods and whether simple structure has been reached.
In some instances we shall see how flaws in the methods have infirmed the results and we shall be able to suggest how the study might be improved.
Naming and interpreting the factors. It is quite possible to carry out technically adequate factor analyses and to fail to interpret the factors correctly. Thus, for example, a factor may be labelled without regard to similar factors which have emerged from other studies. In other cases the fact that a factor is correlated with other factors may be ignored. Even more importantly I shall examine the reasons for interpreting and naming a factor. This process will be most helpful in seeing how factors are interpreted.
