When a search list has been developed which it is considered can be applied reliably to a series of comparable texts, this procedure allows their direct comparison with simply by selecting the file names and specifying the search list and other optional settings to be applied. With a bit of experience, this allows you to speed up the process of comparing larger numbers of texts, without the need to examine each one individually in detail.
The text files selected can be compared by Procrustean Individual Differences Scaling (PINDIS), Individual Differences Scaling (INDSCAL) , Singular Value Decomposition or Correspondence Analysis, by clicking on the corresponding button.
The profiles of occurrences in a series of texts of the items contained in a search list can be compared by Singular Value Decomposition, a metric procedure which considers them as 'scores', which are plotted as vectors of equal length in the configuration of points representing the words or categories. Their content can be read as orthogonal projections on the respective vectors and the differences between the various texts as distances on the circle (or sphere) on which they are projected.
An additional procedure, SUBJSTAT, provides an appropriate arc-distance measure for the analysis of the significance of distances between items in the subject space coordinates produced by these scaling procedures.
Alternatively, the text profiles can be compared using Correspondence Analysis. Here, it is important to notice that comparisons of distances between the individual texts and items in the search list are not normally admissible, although conventionally they are jointly displayed.