PCA analysis - identifying extremes of each factor

I am analysing the size and shape of a particular population of people using 25 different body measurements. I completed a PCA/factor analysis in SPSS which suggested 3 factors.

The next step is to identify "boundary manikins" that represent the extremes of each factor so I can then understand the size/shape of the extremes of this population. The step to do this states "This process was made on applying z-scores to the normalised factor outputs gained from the PCA process for each participant. 95th percentile z-scores were selected in the fashion demonstrated in the Figure (see attached) with the output identifying the Z-scores for each factor for each of the boundary manikins - I'm unsure how how to undertake this in SPSS and exactly what it requires me to do.

Any assistance would be greatly appreciated. Cheers