Validity and feasibility of a tri-axial accelerometer for measuring an upper limb Fitts' task

Abstract

Background: Devices used to display and record Fitts' task performance typically lack portability and accessibility outside the laboratory environment. Touch-based interfaces, while more portable, cannot yield many secondary measures of interest. A portable acceleration- and touch-based equipment package was developed to facilitate the application of the Fitts' task in remote environments. Purpose: To validate a novel portable acceleration and touch-based hardware and software package which can assess both primary and secondary Fitts' task performance measures. Methods: Participants (N=22) set up the equipment and tested themselves remotely with live video support. Protocols from Glazebrook et al. (2015), which measured a violation of Fitts' Law using gold-standard equipment, were replicated. Groups of three closely spaced target placeholders were displayed. Participants touched the one filled-in target placeholder. Two measurement methods, touch and acceleration, were compared using linear regressions performed on movement onset and movement offset. Statistical results from both measurement methods were compared categorically to results from Glazebrook et al. (2015). Results: The findings of Glazebrook et al (2015) were replicated using the novel portable acceleration and touch-based equipment. The touch and acceleration methods of measuring movement onset and offset were significantly correlated (p<0.001). Conclusions: A portable acceleration method of measuring the Fitts' task was validated as: 1) it provided similar results to a touch-based method, but with richer secondary data; and 2) it replicated the findings of Glazebrook et al. (2015), which used gold-standard equipment. An alternate hypothesis concerning the cause of the violation of Fitts' Law is discussed.