Factor analysis

Parent Previous Next


Factor analysis is a statistical technique to identify co-occurrence patterns for dimensions, based on the actual distribution of linguistic features (cf. Biber and Gray, Chapter 21). By means of the amount of shared variance, the factors can be extracted one after the other out of the pool of variation. Thus, the first factor that is extracted includes the maximum amount of shared variance whereas the second factor extracts the maximum of the features that are left over and so on. Consequently, only the first factors will contain a significant amount of shared variance (cf. Biber 1988: 82).

Created with the Personal Edition of HelpNDoc: Write EPub books for the iPad