Subjects performed two-dimensional discrete movements either with a helmet-mounted sight or with a joystick. Fitts' Law was found to be a good predictor of the speed-accuracy tradeoff for both systems. The joystick produced faster movement times than the helmet-mounted sight. For both systems, horizontal and vertical movements were slightly faster than diagonal movements. Two dimensional generalizations of Fitts' Law were discussed in terms of multidimensional scaling. The obtained pattern of movement times was found to be intermediate to the predictions of Euclidean and City-block models of the movement space. Muscle coordination strategies were considered, and a strictly serial coordination model was rejected. A strictly parallel model was also rejected for the helmet-mounted sight, but not for the joystick.

Musically trained subjects tapped three beats with their right hand versus two beats with their left hand in synchrony with two corresponding tones. For independent groups of subjects, the pitch difference of the two tones was either small to encourage an integrated perceptual organization or large to encourage a streamed perceptual organization. Integrated versus parallel motor organization was tested by examining the pattern of covariances among intertap intervals. All subjects exhibited integrated motor organization. An integrated multiplicative hierarchical model of motor organization was superior to a serial chained model and to an independent hierarchical model in describing the pattern of covariances. The subjects who heard tones that encouraged an integrated percept performed with less variability than the subjects who heard tones that encouraged a streamed percept. This superior performance with an integrated motor organization and an integrated rather than a streamed perceptual organization is interpreted as evidence for temporal perceptual-motor compatibility.

Subjects manipulated a control stick to move a cursor displayed on an oscilloscope screen through a distance A to a target of width W. Movement times were found to be a linear function of Fitts' index of difficulty, loga(2A/W). The slope of the Fitts' law relationship was considerably steeper when subjects used a velocity control system rather than a position control system. The intercept of the Fitts' law relationship was higher when a position-plussteadiness criterion was used for determining the end of a movement rather than simply a position criterion. An analysis of the trajectories of the cursor revealed the presence of a series of submovements, and their duration, accuracy, and frequency are systematically related to changes observed in the Fitts' law relationship. Conditions under which Fitts' law may not hold are discussed.Fitts' law states that the time to move a stylus from a home position to a stationary target is proportional to an index of difficulty, logs (2.4/HO (Fitts & Peterson, 1964). A is the distance from the home position to the center of the target, and W

People have remarkable difficulty generating two responses that must follow different temporal sequences, unless the temporal patterns are simply related (e.g., periods in 2:1, 3:1 relation). For example, it is hard to tap to two conflicting rhythms presented concurrently (i.e., a polyrhythm) using the right and left hands (Klapp, 1979), or to tap while articulating a conflicting speech utterance (Klapp, 1981). The present experiments indicate that difficulties in processing conflicting rhythms occur even when people must (a) merely monitor the stimuli and indicate the termination of one rhythmic sequence or (b) tap with a single hand. Responding to polyrhythms is thus difficult even without multiple limb coordination. Furthermore, the difficulty of two-handed tapping to polyrhythms that involve two different tones was found to decrease as the pitch difference between the tones was decreased. This result indicates that the difficulty of rhythmic coordination can be perceptually manipulated in a striking fashion. Polyrhythmic performance thus provides an excellent opportunity for examining possible interactions of perceptual and motor organizations.

Scientific research accesses the past to predict the future. The history of science is often best told by those who have lived it. Our purpose is to provide a brief history of human-automation interaction research, including a review of theories for describing human performance with automated systems, an accounting of automation effects on cognitive performance, a description of the origins of adaptive automation and key developments, and an identification of contemporary methods and issues in operator functional state classification. Based on this history and acknowledgements of the state of the art of human-automaton interaction, future predictions are offered.