Abstract
The perception of control is the experience of being causally responsible for events we observe between our bodies and the environment. Here, we investigated the question of how subjective control during visuomotor tasks is modulated by factors other than objective control. We measured how motor behavior and perceived control were influenced by success rate and the spatial location of attention in a simple video game. In Experiment 1, participants used a joystick to move an object on the screen towards a goal. The participants' objective control was varied across conditions by adding pseudo-random noise to the object's movements. Additionally, on some trials an autopilot of varying strength moved the object towards the goal. After each trial, participants rated their control on a scale from 1 to 9. Even though engaging the autopilot reduced the correlation between the joystick and the object's motions, subjective control ratings were higher when the autopilot was strong. In Experiment 2, participants performed a similar task while attending to an RSVP stream of letters slightly in front of the goal. The letters could lie directly along the straight path to the goal or at various angles to either side. On half the trials, an autopilot influenced the object's movement either towards the goal, or to positions 45 degrees clockwise or counterclockwise from the goal. Results showed the object's average path was biased in the direction of the attended location—participants tended to move the object towards where they were looking. This was true regardless of the degree of control, but was modulated by the autopilot direction. This provides empirical support for the folk wisdom that it is best not to look directly at an obstacle when driving or riding a bike, but rather to attend to the path that avoids the obstacle.