Fitting Trials are simple but important experiments, to make sure that a design fits real moving people. This is what to do:
Use PeopleSize data to help design a rough prototype, mock-up or buck. Often it helps to make it adjustable.
Choose some testers (perhaps simply colleagues, students or family) who are very large or very small in the dimensions which matter for usability. Match your users' age if flexibility is relevant.
Use the Measurements TO percentiles facility
of PeopleSize (access is from the main output dialog) to find out exactly what percentiles these
testers are in all the relevant dimensions. Include stature and weight since
these are fundamental and the data are the most accurate.
If necessary, find extra testers to get exact percentiles, or particular combinations,
for important dimensions. It is a good idea to deliberately have some testers
who are outside your design range, to see what the consequences are.
Have your testers mimic real-life use of your design, and record what difficulties,
clearances, reaches etc. they have. Record what you observe not what they say,
and take formal numerical measurements. You may have to correct some habituated
postures, e.g. splayed knees among tall drivers, but in general avoid coaching
so that you have real-world postures.
Make adjustments, or tailor the design, until the best fit is achieved within your other constraints.
Record what percentiles are accommodated by the improved design, and explicitly consider the consequences for users who cannot be accommodated - safety? discomfort? extra instructions? can you stop them buying/using?
Remember that exclusions accumulate: 95th percentile fit for three different dimensions can exclude up to 14 percent of the intended user group (see Design Percentiles).
You may want to make a record of your usefully-sized testers, for the future (see Export).