Abstract
Although rating scales are still a mainstay of data acquisition in psychology and cognitive science, we do not have ample knowledge about how rating decisions unfold over time. In this exploratory study PC cursor trajectories were recorded and analyzed to probe internal states in rating decisions. Major findings are: 1) the variability of cursor trajectories and response RT are highly correlated suggesting that the trajectory variability is the cause of Inverted-U RT effect. 2) The trajectories consist of quick saccadic-like components that are called “strokes” in this paper. The distribution of strokes differed across tasks as a function of cognitive loads. 3) The shape and speed of tangential velocity of trajectories may reflect participant's internal states, especially when cognitive loads are high. Finally, 4) we can infer the decisional vacillation and hesitation using the trajectories. In particular, rating decisions for middle categories are more susceptible to decisional fluctuation. Several suggestions to psychometric theory are provided.
