Making the impossible possible: Leveraging built-in features for non-intrusive and accurate Apple Screen Time...

Area of Science:

• Digital behavior research
• Human-computer interaction
• Mobile computing

Background:

• Traditional digital behavior studies rely on self-reported data, which suffer from recall bias and lack granularity.
• Android devices offer granular behavior tracking via third-party apps, but similar tools are unavailable on iOS due to platform restrictions.
• Significant differences exist between iOS and Android user populations, necessitating methods for studying Apple device usage.

Purpose of the Study:

• To introduce ASTER, a novel procedure for collecting granular Screen Time data from iOS, iPadOS, and watchOS devices.
• To address the gap in passive sensing of digital behavior on Apple devices.
• To enable researchers to extract comprehensive usage data from Apple devices linked to a single Apple ID.

Main Methods:

• Developed a data donation procedure leveraging the synchronization of Apple Screen Time features.
• Utilized system-level files generated on Mac for Screen Time metrics, containing anonymized usage data from linked devices.
• Created a tool to process these system files into a usable dataset (e.g., JSON) for researchers.

Main Results:

• ASTER enables the extraction of granular app usage data from iPhones, iPads, and Apple Watches.
• The procedure provides insights into cross-device behavior without requiring significant technical expertise or financial investment.
• The processed dataset offers detailed information on user engagement with various applications.

Conclusions:

• ASTER represents a significant advancement in passive sensing for digital media research involving Apple devices.
• The method facilitates the integration of cross-device Apple behavior into digital media research.
• Limitations include the requirement of a Mac, a 4-week data capture window, and vulnerability to Apple's software changes.