Two-Part Models Capture the Impact of Gain on Pointing Performance

  • Authors:
  • Garth Shoemaker;Takayuki Tsukitani;Yoshifumi Kitamura;Kellogg S. Booth

  • Affiliations:
  • University of British Columbia;Osaka University;Tohoku University;University of British Columbia

  • Venue:
  • ACM Transactions on Computer-Human Interaction (TOCHI)
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

We establish that two-part models of pointing performance (Welford’s model) describe pointing on a computer display significantly better than traditional one-part models (Fitts’s Law). We explore the space of pointing models and describe how independent contributions of movement amplitude and target width to pointing time can be captured in a parameter k. Through a reanalysis of data from related work we demonstrate that one-part formulations are fragile in describing pointing performance, and that this fragility is present for various devices and techniques. We show that this same data can be significantly better described using two-part models. Finally, we demonstrate through further analysis of previous work and new experimental data that k increases linearly with gain. Our primary contribution is the demonstration that Fitts’s Law is more limited in applicability than previously appreciated, and that more robust models, such as Welford’s formulation, should be adopted in many cases of practical interest.