Using each athlete's single best-ever performance rather than one time-point measurement.
A peak-selection (or 'Best-Ever') filter draws each participant's single best lifetime result across multiple competitions/attempts, rather than a single standardised test occasion. This can inflate upper-percentile norms relative to what any one athlete would produce on a given day, which matters when such norms are used for single-session clinical benchmarking.
This guide was auto-drafted and is pending editorial review.