A Comparative Study of Elderly, Younger, andChronically Ill Novice PDA UsersKatie A. Moor1, Kay Connelly1, and Yvonne Rogers21Indiana UniversityComputer Science Department{kmoor, connelly}@cs.indiana.edu2Indiana UniversitySchool of [email protected]. Some researchers in the UbiComp community create applica-tions for diverse groups of people. However, before these applications canbe developed, we must ensure the target user group can use the tech-nology. Researchers are challenged to investigate new ways of helpingolder people and chronically ill people remain independent and preservetheir quality of life. Some of these assistive solutions require elderly peo-ple and chronically ill p e ople to use Personal Digital Assistants (PDAs).Critics question whether elderly and chronically ill people can use PDAsgiven their difficulties with computers. This paper presents two usabilitystudies showing there are no major differences in performance between el-derly, chronically ill, and younger people p erforming traditional (pressingbuttons, viewing icons, recording messages) and non-traditional (scan-ning barco des ) P DA tasks.1 IntroductionResearchers in the UbiComp community are creating applications for diversepopulations of people. Some researchers chose to develop ubiquitous applica-tions for elderly people or people with chronic illness to assist people live in-dependent and productive lives. Personal Digital Assistants (PDAs) ([2][3]) andsmart phones ([7]) are some of the devices researchers are using to create assistivetechnologies for elderly and chronically ill p eople. We are currently developing aPDA application to help elderly and younger people who suffer from illnesses re-quiring dietary restrictions (end-stage renal failure, diabetes, etc.) monitor theirnutrition.Since we started our nutrition monitoring project, we have been cautionedthat elderly and chronically ill people may not be able to use PDAs, given theadverse effects age has on vision, dexterity, and coordination [4][5]. If elderly andchronically ill populations have difficulties using computers, how can they usePDAs with smaller screens and buttons?In this paper, we present two of our initial studies investigating if elderly andchronically ill people can use PDAs. In our first study, we conducted a usabilitystudy on two groups of people: 25-30 years old and 75-85 years old. Participantswere asked to completed three traditional PDA tasks (pushing buttons, viewingicons, and recording voice messages) and two non-traditional tas ks (scanningbarcodes with two kinds of scanners). The non-traditional tasks were addedto aid in our research for a future method of inputting nutrition information.Participant performance in scanning can give insight into how participants mayfair in other non-traditional PDA tasks such as taking digital pictures.Our first study found elderly people can complete tasks just as well as people25-30 years old. Elderly people needed a little more practice time to learn how tocomplete each task, but overall there is no difference in performance. In addition,the elderly tended to express slightly less frustration with the technology thanour younger user group. Thus, while the elderly’s performance may be somewhatdegraded because of p oorer vision, coordination, and dexterity, this does nottranslate to an inability or unwillingness to use the technology.Our second study investigated if chronically ill people can use PDAs andcompared the results with our first study. Similar to the first study, participantscompleted three traditional tasks (pushing buttons, viewing icons, and recordingvoice messages) and one non-traditional task (scanning barcodes with the favoredscanner from the first study). We found that chronically ill people can completetasks just as well as younger and elderly people.This paper begins by briefly reviewing related work. The technology, appli-cations, and evaluation techniques we used in our studies are discussed next. Weend with a description of our user studies evaluation of our results, and ideas forfuture work.2 Related WorkUsability studies have been conducted on many PDA applications. However thestudies usually focus on the interface design and presentation of information,instead of the user’s ability to use the PDA. In this section, we will brieflydescribe some research that has been done with respect to icon sizes and voicerecording.Stephen Brewster evaluated how multi-modal feedback can help improve dataentry in PDA computers regardless of icon size. Multi-modal feedback involvesthe PDA device making sounds, tactile changes (i.e. vibration) or vision changes(button turns to inverse colors when touched) when PDA tasks are completed.He evaluated how college students performed inputting numbers into a PDAwhile sitting and walking. The experiments showed sound helped users inputmore numbers, however smaller buttons required more “workload” [1].Using PDAs for voice recordings is becoming a popular way to get userfeedback in situ. Stephen Intille et al. integrated voice recordings into theircontext-aware experience sampling tool to obtain feedback from participants [8].In the future, we see researchers using the PDA audio recorder for voice diariesto allow users to quickly record problems or comments about applications formobile usability testing.To the best of our knowledge, there have not been any extensive studiescomparing the abilities of different age groups and people with chronic illnessusing PDAs.3 General Exper iment DesignBefore we began the usability study, we had to evaluate what PDAs and scannersto use in the study. Once we decided on the hardware, we looked into the abilitieswe wanted to test and how to create applications to test our users abilities. Afterwe created our applications, we decided how to measure the usability of thePDA. In this section, we will discuss how we selected the Tungsten T3 as ourPDA, Socket SDIO scanner and Baracoda pencil as our scanners, and a modifiedversion of NASA’s Task Load Index (TLX) as our metric for our usability testingapplications.3.1 HardwarePDA Selection The PDA is the center of our study and the building block ofmany UbiComp applications, thus we had to evaluate current PDAs available.Most UbiComp PDA applications use off-the-shelf PDAs to make their applica-tion more widely accessible and less expensive. We conducted our study with anoff-the-shelf PDA so the results would be useful to the