| | 65 | Comma Separated Value files can be imported into most spreadsheets and statistics packages. Each row in the csv file stores that data for a particular face, each column the parameters of a particular component. The columns are in order, most variance explained (left) to least variance explained (right). The rows can appear confusing, as unfortunately Psychomorph does not output the name of the template file used to generate the parameters. The faces are analysed and the parameters generated in the exact order they appear in the Batch file supplied to the ‘'''Batch Analyse Shape'''’ menu. To you can simply take a batch file, load it into you stats or spreadsheet program (most have an import text or csv function) and each name in the batch file will exactly match, row for row, the parameters output. This can be a little disconcerting, but you can reconstruct the parameters to check. |
| | 67 | To generate the Masculinity scores (the whole point of the tutorial if you can remember that far back) we used Linear Discriminant Analysis. This was calculated in SPSS. A sex variable (1- male, 0-female) was defined and used as the ''grouping variable''. The order isn't important, you can define 1-female, 0-male if it makes you feel better, just remember which way round you did it :). The parameters from the ASM model you generated in Psychomorph are the ''Independents''. In order to generate a Masculinity score we asked SPSS to save the Discriminant Scores, unlike the probabilities of group membership these scores are linear. |
