Title: Application and device effects on the usability of mobile phones and PDAs
Submission type: Research paper
Relevance to magazine:
Presentation and clarity:
Depth of contribution:
Detailed comments for the authors:
This paper presents a usability comparison of two approaches for cross-platform deployment on mobile devices. 2 pilot studies examine the usability of applications that are automatically ported between mobile devices (phones and PDAs) by j2me and the usability of an application that is tuned to the constraints of a particular device (PDA). Results show that benefits exist for manually porting applications to mobile devices that may outweigh the efficiency benefits of automation.
This paper provides a nice statistical breakdown of task efficiency and user preference for applications that have been ported from the PDA to the phone using j2me. However, the authors provide few details on the layout of the applications themselves. In addition, the details of how j2me performs the porting are left out. How is stylus input on the PDA supplanted by keypad input on the phone? How are widgets scaled, substituted, or divided to accomodate the phone's limited screen space? How are colors substituted? Without these details, it is difficult to understand the results presented for the first pilot study as well as the comparison of both studies at the end of the paper.
The second pilot study lacks any quantitative measures, and is based on observation and informal interviews with 5 expert users of the (desktop) system. While this section of the paper brings up some good issues (handedness, text input using the ipaq's 'pen mode'), the work seems only the first stage in an exercise of iterative design for an application that differs from those compared in the first pilot study. For concrete results, wouldn't the authors need to take one of the applications from the first study and tune it to a phone/PDA, making use of the issues outlined in section 3.2? Due to the informality of this study, the authors are left to only hypothesize that advantages can be gained from developing specifically for the target mobile device. It seems to me that this is fairly obvious from years of research in UIMS and more recent work in model-based UI techniques.
A few papers recommended:
Myers et al "past present future of UI software tools" , TOCHI 7(1), march 2001.
discusses the failures of UIMS, a nice historical lesson that applies to the concepts addressed in this paper.
Eisenstein et al, "applying model-based techniques to the development of ui's for mobile computers", IUI'2001
discusses an architecture for intelligently porting UI's across mobile devices. Provides some references for techniques that can be used to substitute or scale widgets, adjust layout, and adjust color when porting to a particular device.