Bias in Experimental Comparisons between Equipments due to the Order of Testing

Abstract

Trials comparing equipments should use separate groups of people for each equipment, If the same people work with all the equipments in balanced orders, the results of the trial may be biased by hidden transfer effects. This point is illustrated by two sets of experiments.

One set of experiments compared true motion or pursuit displays with relative motion or compensatory displays. True motion displays are always preferable to relative motion displays. Yet three experiments found a relative motion display reliably better than a true motion display under certain conditions. Those unfortunate results are probably due to the people in the experiments confusing the various optimal phase relationships between control movements and display movements.

The second sot of experiments compared various orders of control using a true motion display. A position control system is more compatible with a true motion display than is any higher order of control system. Yet one experiment found rate and rate-aided control systems reliably better than a position control system with the lowest frequency track. This result also is probably due to confusion between the various optimal phase relationships.