Nonadjectival Rating Scales in Human Response Experiments

Abstract

Twenty-two subjects participated in two tracking experiments for the purpose of determining the utility of a nonordinal, nonadjectival rating scale. The scale was devised in an effort to allow a human to quantify his subjective opinions of the characteristics of a system in situations where an adjectival scale would be inappropriate. The tracking task in both experiments was a compensatory one in which the human operator attempted to minimize the difference between a random-appearing input signal and the output of an unstable, controlled element. The system dynamics and input signal were mechanized on an analog computer. The error signal was viewed by the operator on an oscilloscope screen. Control was effected by a small isometric manipulator. In the first experiment, ratings were generated by changing the degree of instability of the controlled element. In the second, the manipulator sensitivity was varied. The nonadjectival rating concept shows definite potential for use in a wide variety of situations in which human opinion is elicited.