HardBorders: a new haptic approach for selection tasks in 3D menus

Proceedings of the 16th ACM Symposium on Virtual Reality Software and Technology - VRST '09 (2009), pp. 243-244, doi: 10.1145/1643928.1643986
Experiments
Tasks
Findings

How do different haptic modalities (none, magnetic, HardBorders) affect the time taken and accuracy of selecting items in a menu?

Variables and Constants

    Haptic Fidelity
  • haptic quality - none, magnetic attraction around target, hardborders impassable border
Constants

Number of items in menu: 8; Diameter of the circle around the menu: 8 cm; Inclination of the plane: 20 degrees.

System Info

Displays
  • monitor - 17 inch 2d screen, 1280x1024 resolution
Input Hardware
  • haptic - phantom premium 1.5
Software
None

Participant Info

The participants had different levels of experience with haptics interfaces.

Total # Age Range Gender Balance
24 -

The participant pressed a button on the PHANToM device to show the menu centered at the pointer's location, and then selected one of the 8 menu items.

Interaction and Environment

Interface

The users had a training session beforehand. The user held the PHANToM device pen, directed it to a location, and pressed a button to show the menu centered at that location. The user could then use the pen to direct the pointer to the correct menu item. There were 20 selection tasks per participant, with each participant taking a total of 10 minutes for the experiment.

Dimensionality Scale Density Visual Realism
2.5D Small Low Low
Metrics

  • time - time taken from showing the menu by pressing the button to selecting the menu item.
  • accuracy - precision in mm for multiple selections, number of target re-entries, and number of axis crossings.
  1. There was a significant inverse effect of haptic quality on time for a object selection task.

    HardBorders took the least time, with the magnetic technique second, and no haptics the slowest.

    Specificity: Somewhat specific

  2. There was a significant direct effect of haptic quality on accuracy for a object selection task.

    HardBorders provided the highest precision gain, with the magnetic technique second, and no haptics the lowest.

    Specificity: Somewhat specific