A Large-Scale Study of iPhone App Launch Behaviour

Abstract

There have been many large-scale investigations of users' mobile app launch behaviour, but all have been conducted on Android, even though recent reports suggest iPhones account for a third of all smartphones in use. We report on the first large-scale analysis of app usage patterns on iPhones. We conduct a reproduction study with a cohort of over 10,000 jailbroken iPhone users, reproducing several studies previously conducted on Android devices. We find some differences, but also significant similarities: e.g. communications apps are the most used on both platforms; similar patterns are apparent of few apps being very popular but there existing a 'long tail' of many apps used by the population; users show similar patterns of 'micro-usage'; almost identical proportions of people use a unique combination of apps. Such similarities add confidence but also specificity about claims of consistency across smartphones. As well as presenting our findings, we discuss issues involved in reproducing studies across platforms.

App Usage Predicts Cognitive Ability in Older Adults

“…However, Social Networking apps make up only half as many of launches for older adults as they do for younger adults, at 6 App usage by hour of day Figure 1: Older adults use their phones earlier in the day than younger adults. We show app launches by hour of day from our dataset of older adults overlayed on the results reported by Morrison et al [36] for younger adults. The y-axis is normalized as percent of total usage to allow for comparison between populations with different sample sizes.…”

“…The second unifying factor is that, as pointed out by Morrison et al [36], nearly all of this work has been performed using Android phones rather than on iPhone and iOS, as the Android platform allows for the easy installation of logging software for OS-level events such as app launches and phone unlocking. Morrison responded by reproducing on iPhone a variety of studies previously performed on Android.…”

“…One difficulty with this comparison is that, as noted by Morrison, users of jailbroken devices likely differ from the general population in that they may be more engaged and familiar with their iPhones. Therefore, their usage patterns can likely be considered on the more 'extreme' end of what we may expect from the general population [36].…”

“…We begin by reproducing analyses by Morrison et al [36]. As mentioned in our Prior Work section, this is the only other paper to have characterized iPhone app-usage statistics.…”

Student Perspectives on Digital Phenotyping

“…The work we are describing does have strong parallels with other forms of smartphone sensor-based study, particularly smartphone usage analytics [12,23,30,57,88], and a growing body of work on occupational stress (e.g. [29,51]).…”