Mixing Board Versus Mouse Interaction In Value Adjustment Tasks (1110.2520v1)

Abstract: We present a controlled, quantitative study with 12 participants comparing interaction with a haptically enhanced mixing board against interaction with a mouse in an abstract task that is motivated by several practical parameter space exploration settings. The study participants received 24 sets of one to eight integer values between 0 and 127, which they had to match by making adjustments with physical or graphical sliders. Based on recorded slider motion path data, we developed an analysis algorithm that identifies and measures different types of activity intervals, including error time moving irrelevant sliders and end time in breaks after completing each trial item. Our results showed a significant increase in speed of the mixing board interaction accompanied by reduced perceived cognitive load when compared with the traditional mouse-based GUI interaction. The gains in speed are largely due to the improved times required for the hand to reach for the first slider (acquisition time) and also when moving between different ones, while the actual time spent manipulating relevant sliders is very similar for either input device. These results agree strongly with qualitative predictions from Fitts' Law that the larger targets afforded by the mixer handles contributed to its faster performance. For further investigation, we computed a measure of motion simultaneity based on velocity correlation, which allowed to identify types of items for which increased simultaneous adjustments occur. For continuous parameter space exploration our findings suggest that mixing boards are a good option to provide detailed multi-value control. The strengths of this input method particularly show in settings where screen space is precious and undisrupted visual focus is crucial.