The MRSCAL algorithm is a metric counterpart to MINISSA. Its aim is to position a set of stimulus objects as a set of points in a space of minimum dimensionality in much the same way as MINISSA, except that the distances in this space will be a linear (or optionally a logarithmic) function of the dissimilarities between the stimuli. It is included here primarily as an alternative scaling procedure to try in the event that MINISSA produces a so-called 'degenerate solution', in which the fitted values all have the same value or 'collapse' into a very few groups.
It takes as input a matrix of standardised joint frequencies which is generated by HAMLET joint frequencies
It is possible to repeat scaling for different combinations of items included in the original word list used by HAMLET without the need to generate the matrix each time.
The resulting configuration is automatically displayed in graphic form : the first two reference axes appear as a horizontal plane on to which labelled points are projected corresponding to words in the text file originally subject to analysis. The configuration can be rotated and zoomed for closer examination, printed or saved for inclusion in other documents. Note that, for easier viewing, configurations are rescaled in these displays by their largest absolute coordinate value.
You can examine and edit the file containing the detailed configuration data on which the display is based.
The results of applying the same vocabulary to a series of different texts can be saved for separate comparison using Procrustean Individual Differences Scaling (PINDIS) . For this purpose, the results of each MDS scaling must first have been saved. PINDIS has an option to call SELECT to choose the configurations to be compared. SELECT can also be used independently to create an input file for PINDIS. If the matrices created for a series of texts have been saved, it is also possible to compare these, using a version of INDividual differences SCALing (INDSCAL).