Use of Mobile Appointment Scheduling Devices Thad E. Starner, Cornelis M. Snoeck, Benjamin A. Wong, and R. Martin McGuire College of Computing and GVU Center Georgia Institute of Technology Atlanta, GA 30332-0280 USA {thad,cmsnoeck,bbb,haileris}@cc.gatech.edu ABSTRACT One hundred thirty-eight subjects participated in a study on mobile appointment scheduling. Subjects completed a ques- tionnaire on their primary method of managing appointments when away from their desks. Immediately afterwards, sub- jects completed a session of scheduling four appointments with the interviewer. The most common scheduling systems, in order of popularity, were paper-based day planners, mem- ory, scrap paper, and PDA’s. However, 43% of the claimed PDA users and 68% of day planner users switched to an- other, more readily accessible method when scheduling an appointment. Interviews revealed a practice of using mem- ory or scrap paper to “buffer” appointments for later entry into the PDA or planner. Categories & Subject Descriptors H.5.2 Handheld Devices and Mobile Computing Keywords mobile computing, PDA, calendar, scheduling INTRODUCTION The personal digital assistant (PDA) market has now reached the level of billions of dollars a year. Yet informal surveys of technical audiences reveals that only a third to one half of PDA owners carry their device. Here, we study one of the most common uses of a PDA, appointment scheduling. We show that both PDA and paper-based day planner users demonstrate a reluctance to use their devices, preferring to revert to memory or scrap paper. We hypothesize that the length of time required to access a scheduling method signif- icantly effects how often it is used. We perform timing tests to support this hypothesis. MOBILE APPOINTMENT SCHEDULING USER STUDY To gain insight into currently used scheduling strategies, we performed a user study inside the main entrance of Georgia Tech’s Student Center, asking passersby to volunteer as sub- jects. A total of 138 subjects participated in the study, with a predominance of young male students (88% age 18-25, 70% male, 90% students). Copyright is held by the author/owner(s). CHI 2004, April 24–29, 2004, Vienna, Austria. ACM 1-58113-703-6/04/0004. The study consisted of two parts: a short questionnaire and an appointment scheduling session. The questionnaire re- quested demographic information, an exhaustive list of all calendar systems used by the subject, the primary calendar system the subject uses when away from his desk, how long the subject has been using this calendar system, and how many appointments the subject schedules per week. Eight Likert scale questions (ranging from 1 for strong agreement to 7 for strong disagreement) were used to elicit the subjects’ opinions on their calendar systems (ease of use, necessity, adequacy, expense, etc.). We ended the survey with open- ended questions on the most liked features of the subject’s current calendar system. Next, each participant was asked to sit at a table for an inter- view with one of our researchers to perform timing tests on appointment scheduling practices. A single researcher per- formed all of the appointment scheduling with the subjects, using a script of four tasks: A. Schedule an appointment for a date seven days in the fu- ture. ‘Could we meet sometime next Monday?”. B. Schedule an appointment for a date three months in the future. ‘Could we schedule a time to meet in the second week of February?”. C. Schedule an appointment for tomorrow. ‘Could we sched- ule a time to meet tomorrow?”. D. Reschedule the second appointment to the next day. ‘Could we reschedule our appointment in February from the 10th to the 11th?”. Subjects were encouraged to schedule the appointments as if they were of significant importance, and scheduling con- flicts were resolved as part of the task but not included in the recorded times. In order to capture timing data accurately, the experiment was videotaped with two cameras: one pointing forward to- ward the test subject and a second pointing down at where test subject was likely to place their scheduling device while in use (see Figures 1 and 2). The cameras were time synchro- nized, and both recorded audio. CHI 2004 ׀Late Breaking Results Paper 24-29 April ׀Vienna, Austria 1501