Repertory grid analysis for probation officers

In order to investigate the relatedness of the constructs and to find if there was any difference in the patterning of the constructs between the mentally disordered clients and the controls, factor analysis was employed. The analysis was carried on the standardized and unstandardized constructs in order to investigate the scaling problem. Two probation officers had all the 24 constructs in which the within probation officer standard deviation was zero so that standardization was impossible, hence they were excluded from the analysis. Thus the analysis was carried out on the remaining 140 pairs. Five probation officers, each with several mentally disordered clients had one or more constructs with a nil standard deviation. In such a case a standardized Z score of zero was computed for the constructs. In the analysis of the unstandardized constructs, the constructs with zero standard deviation were excluded by computing zero score for them. Hence we had an equal number of clients excluded from both types of analyses.