Movement Time Model Research Articles

Movement and positioning of a virtual cuboid in 3d space in an augmented reality environment

Human-virtual object interaction is common in both entertainment and work settings. This study investigated the movement time (MT) and subjective rating of difficulty (SRD) for moving a virtual cuboid in a 3D space. The participants wore an augmented reality (AR) headset, picked up a virtual cuboid, and placed it on an assigned target. They rated the SRD of the task on a five-point scale. The effects of the 3D coordinate of the target, sex, and handedness on the MT were analysed. The error placement rate was also recorded. Significant effects of spatial coordinates were found on both MT and SRD. Both single- and two-stage MT modelling were conducted using segmented and unsegmented MT data, respectively. The insignificant prediction error between the models indicates that the two-stage MT model is not superior to the single-stage one. The findings of this study are beneficial to software designers in designing user-friendly AR applications.

Modeling the Effect of Target Shape on Movement Performance in a 1D2D Fitts Task

Fitts’ law is used as a performance measurement metric in human–computer interactions. The original formulation implied that movement time was identical for movements with the same value of the index of difficulty under varied movement amplitude and target width. An experiment was designed to test this implication. The result indicates that movement time is related to the index of difficulty when the amplitude is constant. Nowadays, most of the icons in applications are represented as two-dimensional targets. An object of equal width and height is a particular case of a two-dimensional target. This target area could be a factor in a Fitts task and impact the movement time, number of errors, and perceived difficulty. Therefore, the area could replace the target width in the formulation of the index of difficulty. The modified index of difficulty is easy to implement without the complexity of post-calculation. Researchers can design the index of difficulty before the empirical test. This research proposes a modified index of difficulty by varying the target’s area and applying the square-root movement time model simultaneously, which results in an excellent performance with a higher R-square and satisfies the residual normality robustly than the traditional formulation of Fitts’ law.

Human-centric predictive model of task difficulty for human-in-the-loop control tasks.

Quantitatively measuring the difficulty of a manipulation task in human-in-the-loop control systems is ill-defined. Currently, systems are typically evaluated through task-specific performance measures and post-experiment user surveys; however, these methods do not capture the real-time experience of human users. In this study, we propose to analyze and predict the difficulty of a bivariate pointing task, with a haptic device interface, using human-centric measurement data in terms of cognition, physical effort, and motion kinematics. Noninvasive sensors were used to record the multimodal response of human user for 14 subjects performing the task. A data-driven approach for predicting task difficulty was implemented based on several task-independent metrics. We compare four possible models for predicting task difficulty to evaluated the roles of the various types of metrics, including: (I) a movement time model, (II) a fusion model using both physiological and kinematic metrics, (III) a model only with kinematic metrics, and (IV) a model only with physiological metrics. The results show significant correlation between task difficulty and the user sensorimotor response. The fusion model, integrating user physiology and motion kinematics, provided the best estimate of task difficulty (R2 = 0.927), followed by a model using only kinematic metrics (R2 = 0.921). Both models were better predictors of task difficulty than the movement time model (R2 = 0.847), derived from Fitt’s law, a well studied difficulty model for human psychomotor control.

A model for combined targeting and tracking tasks in computer applications

Current models for targeted-tracking are discussed and shown to be inadequate as a means of understanding the combined task of tracking, as in the Drury's paradigm, and having a final target to be aimed at, as in the Fitts' paradigm. It is shown that the task has to be split into components that are, in general, performed sequentially and have a movement time component dependent on the difficulty of the individual component of the task. In some cases, the task time may be controlled by the Fitts' task difficulty, and in others, it may be dominated by the Drury's task difficulty. Based on an experiment carried out that captured movement time in combinations of visually controlled and ballistic movements, a model for movement time in targeted-tracking was developed.

Verification of models for ballistic movement time and endpoint variability

A hand control movement is composed of several ballistic movements. The time required in performing a ballistic movement and its endpoint variability are two important properties in developing movement models. The purpose of this study was to test potential models for predicting these two properties. Twelve participants conducted ballistic movements of specific amplitudes using a drawing tablet. The measured data of movement time and endpoint variability were then used to verify the models. This study was successful with Hoffmann and Gan's movement time model (Hoffmann, 1981; Gan and Hoffmann 1988) predicting more than 90.7% data variance for 84 individual measurements. A new theoretically developed ballistic movement variability model, proved to be better than Howarth, Beggs, and Bowden's (1971) model, predicting on average 84.8% of stopping-variable error and 88.3% of aiming-variable errors. These two validated models will help build solid theoretical movement models and evaluate input devices. Practitioner summary: This article provides better models for predicting end accuracy and movement time of ballistic movements that are desirable in rapid aiming tasks, such as keying in numbers on a smart phone. The models allow better design of aiming tasks, for example button sizes on mobile phones for different user populations.

Improving on Human’s Movement Time Model for Pointing Task - Introducing the Target Position Factor into Fitts’ Law

The experiments were aimed at determining the effect of target position on movement time when performing a target pointing movement task. 21 subjects performed pointing movements with 72 conditions of various target position. The movement time data were collected. It was shown that the starting point position and target position greatly affect the movement time. As a result, the movement times were not explained satisfactorily by the conventional Fitts’ model. The conventional model was improved by introducing the target position factor into. Compared with the conventional Fitts’ model, the new model could describe the data better, both in term of contribution value (r2) and the standard error of the residual between the predicted value by model fit and the measured movement time.

Bivariate pointing movements on large touch screens: Investigating the validity of a refined Fitts’ Law

On the basis of three empirical studies Fitts' Law was refined for bivariate pointing tasks on large touch screens. In the first study different target width parameters were investigated. The second study considered the effect of the motion angle. Based on the results of the two studies a refined model for movement time in human-computer interaction was formulated. A third study, which is described here in detail, concerns the validation of the refined model. For the validation study 20 subjects had to execute a bivariate pointing task on a large touch screen. In the experimental task 250 rectangular target objects were displayed at a randomly chosen position on the screen covering a broad range of ID values (ID= [1.01; 4.88]). Compared to existing refinements of Fitts' Law, the new model shows highest predictive validity. A promising field of application of the model is the ergonomic design and evaluation of project management software. By using the refined model, software designers can calculate a priori the appropriate angular position and the size of buttons, menus or icons.

Ballistic movements on data-entry keypads

Research to date on movement times on data-entry keypads has assumed that the movements are made under visual control [Drury, C.G., Hoffmann, E.R., 1992. A model for movement times on data-entry keyboards. Ergonomics 35, 129–147; Hoffmann, E.R., Tsang, K.T., Mu, A., 1995. Data-entry keyboard geometry and keying movement times. Ergonomics 38 (5), 940–950; Silfverberg, M., MacKenzie, I.S., Korhonen, P., 2000. Predicting text entry speed on mobile phones. In: Proceedings of the ACM Conference on Human Factors in Computing Systems – CHI 2000. ACM, New York, pp. 9–16]. This is often not the case; where the value of Fitts’ Index of Difficulty is low, it is likely that movements will be made ballistically. A survey of commonly used keypads on devices such as mobile phones and ATMs shows that there is a need to study movement times when these ID values are less than those for which movements are likely to be made under visual control [Gan, K.-C., Hoffmann, E.R., 1988a. Geometrical conditions for ballistic and visually-controlled movements. Ergonomics 31, 829–839]. A series of experiments on simulated and real keypads indicates that the ballistic form of movement is generally valid and may be modelled by a modification of the ballistic movement time model of Hoffmann [1981. An ergonomics approach to predetermined motion time systems. In: Australian Institute of Industrial Engineers, Proceedings, Ninth National Conference, pp. 33–47] and of Gan and Hoffmann [1988a. Geometrical conditions for ballistic and visually-controlled movements. Ergonomics 31, 829–839l; 1988b. Sequential ballistic movement. Ergonomics 31 (10), 1421–1436]. A model is developed for the time to complete sequences of number entries containing up to 5 numbers. In this model, the movement times are related to the square-root of the sum of the movement amplitudes. Relevance to industry Many industrial applications require the use of keypads for data entry. Movements may be made ballistically or visually controlled, depending on the geometry of the keypad. As the time for data entry is dependent on geometry, a model for time taken is necessary for predicting human performance.

A model for movement time on data-entry keyboards

A model is presented for the effects of keyboard geometrical design on the movement time between keys, when the movements are carried out under visual control with single finger operation (such as on a data-entry keypad). It is shown that, for a given key centre spacing, the optimum key width occurs when the gap between adjacent keys is equal to the width of the finger pad (that part of the finger which is used for key actuation). Three experiments are reported related to keyboard design. In an experiment on. simulated keyboards and using five different probes (four metal probes of different widths and real fingers), it is found that results similar to the theoretical predictions are obtained in that a minimum in movement time occurs when the inter-key gap is about equal to the probe width. The effect is somewhat less with real fingers when compared with metal probes. In the second experiment, five real keyboards were tested with constant key spacing and with variable key size. A number of different amplitudes were used in this test to cover the range used in practice; thus movements were made both ballistically and under visual control in the different conditions. It was found that the major features of the model were supported in this experiment, although there were differences both within and between subjects. The optimum key size, at which both movement time and error rate were minimized, was about 6 to 8 mm (this is applicable only to the standard key spacing of 19 mm used in the experiment). In this experiment, the effect sizes were generally smaller than those found with metal probes, possibly due to the variation of finger size in the subjects or to adaptability of the human to the different experimental conditions. In the third experiment, reciprocal movements on a calculator keyboard showed good agreement with Pitts' Law when the effective target width, as defined in the paper, was used for calculation of the Index of Difficulty.

A second-order model for movement time.

A second-order model which takes into account both the main and the interaction effects of the independent variables for movement time was developed using data of Fitts and Petersen (1964). The model shows that movement time is significantly affected by the index of movement difficulty and nonsignificantly affected by stimulus information. In addition, the interaction effect is negligible.