Investigating the effects of encumbrance on one- and two- handed interactions with mobile devices

Abstract

In this paper, we investigate the effects of encumbrance (carrying typical objects such as shopping bags during interaction) and walking on target acquisition on a touchscreen mobile phone. Users often hold objects and use mobile devices at the same time and we examined the impact encumbrance has on one-and two-handed interactions. Three common input postures were evaluated: twohanded index finger, one-handed preferred thumb and twohanded both thumbs, to assess the effects on performance of carrying a bag in each hand while walking. The results showed a significant decrease in targeting performance when users were encumbered. For example, input accuracy dropped to 48.1% for targeting with the index finger when encumbered, while targeting error using the preferred thumb to input was 4.2mm, an increase of 40% compared to unencumbered input. We also introduce a new method to evaluate the user's preferred walking speed when interacting -PWS&I, and suggest future studies should use this to get a more accurate measure of the user's input performance.

Comparing evaluation methods for encumbrance and walking on interaction with touchscreen mobile devices

“…More recently, Ng et al [22] assessed the impact of encumbrance on one-and two-handed interactions. Users performed a target acquisition task in three common input postures (two-handed index finger, one-handed preferred thumb and two-handed both thumbs) while walking and carrying multiple objects (a bag in each hand).…”

“…Hyman et al [10] reported that users walked slower and were more unaware to nearby events when using mobile phones. Users could be further affected if they also had to carry bulky objects during interaction, as shown by [21,22].…”

“…In the outdoor mobile study by Kane et al [12], all users walked at the same pre-set pace by following a experimenter and kept within a certain distance. We used a similar method to Ng et al [22] to control walking speed in our experiment as users walked at their own measured PWS.…”

“…Users may struggle to input accurately or in an effective manner when their hands are physically hampered by holding objects at the same time which can cause interaction to become frustrating and prone to errors. As a result, it is important to assess the impact of encumbrance, especially when the user is also on the move, a situation already known to cause interaction problems [21,22]. A greater understanding of the issues caused by encumbrance while walking would mean better interaction techniques and more efficient interfaces can be developed to assist the user in these physically awkward and mentally demanding situations.…”

“…The treadmill guides the user to walk at a constant pace without having to divide attentional resources between interaction, navigation and avoiding nearby obstacles. For walking around a predefined route, we used the method of Ng et al [22] to control the user's PWS by using a human pacesetter. An experimenter who acted as the pacesetter walked at each user's PWS (which was measured at the beginning of the study).…”