An experimental study on the role of touch in shared virtual environments

ACM Transactions on Computer-Human Interaction (2000), pp. 443-460, doi: 10.1145/365058.365082
Experiments
Tasks
Findings

What is the effect of haptics on task performance and the sense of "togetherness" in a shared virtual environment?

Variables and Constants

    Haptic Fidelity
  • haptics (on/off) - haptic force-feedback through phantom device or none
Independent Variables
  • order - half of the participants used the haptic condition first, and the other half the non-haptic condition first
Constants

Graphics update rate: 30 Hz; haptic update rate: 1000 Hz

System Info

Displays
  • monitor - two approximately 20" monitors
Input Hardware
  • haptic - two phantom haptic devices
Software
None

Open Inventor graphics toolkit; single master computer with bifurcated video signal to show each user the same virtual environment.

Participant Info

Total # Age Range Gender Balance
10 -

Users moved their PHANToM device pens to move their cursors, each attached to a ring around a piece of curved wire. The ring had an inner diameter several times larger than the wire's diameter. Each user could apply force to the ring by moving their pens, with both users needing to contact the ring and apply simultaneous lateral force in order to move it. The task was to move the ring along the wire from one end to the other with minimal contact to the wire.

Interaction and Environment

Interface

Users moved their cursors by moving their PHANToM device pen,. though these cursors were attached to the ring at a diameter of the ring. The ring could be held by applying force on each cursor, one from each user. It could then be moved through both users applying simultaneous lateral force. The ring did not fall down if it was not held. If force was not applied by both users at the same time, or below a force threshold, the ring did not move. One of the two users was the same expert user, a confederate, for all trials, though the participant did not know this. The expert performed as consistently as possible across all trials. Half of the users used the haptic condition first, waited ten days, and then used the non-haptic condition. The other half did the opposite.

The environment consisted of one wall and a floor with the wire in the air above the floor. The ring started around the wire at one end. If the ring contacted the wire, the walls and ring changed colors until the ring was moved back out of contact with the wire.

Dimensionality Scale Density Visual Realism
3D Small Low Low
Metrics

  • presence - 8 questions on togetherness embedded in a larger general questionnaire administered after the experiment
  • accuracy - metric equal to (a*r^2)/t, where a is a constant, r the ratio of time spent in error-free condtion to total time, and t the total time
  1. There was a significant interaction between haptics (on/off) and order on accuracy for a object manipulation task.

    Users who experienced the non-haptic condition first and then the haptic system second performed better on the metric than those with the opposite ordering. The authors suggest this has to do with training on the system, which was then enhanced with the a

    Specificity: Somewhat general

  2. There was a significant direct effect of haptics (on/off) on presence for a object manipulation task.

    Users reported their sense of togetherness to be higher in the haptic condition than in the non-haptic condition. When linked with collected demographic information, togetherness decreased with age, and females reported higher levels of togetherness than

    Specificity: Highly general

  3. There was a significant direct effect of haptics (on/off) on accuracy for a object manipulation task.

    Users performed better by the metric used by the authors in the haptic condition than in the non-haptic condition.

    Specificity: Highly general