The behavioral illusion : the misperception of voluntary behavior
Jordon, Jerome Scott
MetadataShow full item record
The present study investigated what William T. Powers, in 1978, called the Behavioral Illusion: When an individual controls the value of an environmental variable by monitoring that variable and offsetting would be disturbances with compensatory output, the compensatory output often appears to constitute the individual's behavior, even though that output is driven by the disturbance. Conversely, the controlled input often goes unrecognized as the individual's behavior. Two experiments were conducted to examine the phenomenon. In Experiment 1, subjects passively watched for three minutes as an operator controlled the horizontal position of a symbol (x) on a computer screen by offsetting computer-generated disturbances with a computer mouse. The mouse's movements were reflected by another symbol (o) on the screen. Each subject observed the operator either continuously or discontinuously engaged in the task, with 80 subjects in each condition. In the discontinuous condition, the operator released the mouse during the second minute. At the end of the three minutes, subjects answered the question "Which of the two letters best represents the operator's behavior, x or o?" Experiment 2 differed from Experiment 1 only in that the subjects (80, 40 in each condition) actively manipulated a computer joystick which influenced the position of either x or o, depending upon a subject-determined switch setting. Subjects in the continuous and discontinuous conditions selected the symbol o with a relative frequency of 87.5% and 80%, respectively, in Experiment 1, and 70% and 52.5%, respectively, in Experiment 2. A multidimensional X² analysis of the combined data yielded a significant main effect for each independent variable, with no interaction: continuous vs. discontinuous; X² (1) = 10.88, p < .001 and, active vs. passive; X² (1) = 8.98, p < .001. Being unable to see the computer disturbance which drove the compensatory mouse movements reflected by the symbol o, subjects tended to perceive the position of symbol o as being the operator's, rather than the disturbance's, doing. However, the incidence of this misperception was reduced when subjects viewed discontinuities in the behaver's conduct, or worked at cross-purposes with the behaver. Implications for attribution theory are discussed.