Systems for Managing Work-Related Transitions

Abstract

Peoples' work lives have become ever-populated with transitions across tasks, devices, and environments. Despite their ubiquitous nature, managing transitions across these three domains has remained a significant challenge. Current systems and interfaces for managing transitions have explored approaches that allow users to track work-related information or automatically capture or infer context, but do little to support user autonomy at its fullest.

In this dissertation, we present three studies that support the goal of designing and understanding systems for managing work-related transitions. Our inquiry is motivated by the notion that people lack the ability to continue or discontinue their work at the level they wish to do so. We scope our research to information work settings, and we use our three studies to generate novel insights about how empowering peoples' ability to engage with their work can mitigate the challenges of managing work-related transitions.

We first introduce and study Mercury, a system that mitigates programmers' challenges in transitioning across devices and environments by enabling their ability to continue work on-the-go. Mercury orchestrates programmers' work practices by providing them with a series of auto-generated microtasks on their mobile device based on the current state of their source code. Tasks in Mercury are designed so that they can be completed quickly without the need for additional context, making them suitable to address during brief moments of downtime. When users complete microtasks on-the-go, Mercury calculates file changes and integrates them into the user's codebase to support task resumption.

We then introduce SwitchBot, a conversational system that mitigates the challenges in discontinuing work during the transition between home and the workplace. SwitchBot's design philosophy is centered on assisting information workers in detaching from and reattaching with their work through brief conversations before the start and end of the workday. By design, SwitchBot's detachment and reattachment dialogues inquire about users' task-related goals or user's emotion-related goals. We evaluated SwitchBot with an emphasis on understanding how the system and its two dialogues uniquely affected information workers' ability to detach from and later reattach with their work.

Following our study of Mercury and SwitchBot, we present findings from an interview study with crowdworkers aimed at understanding the work-related transitions they experience in their work practice from the perspective of tools. We characterize the tooling observed in crowdworkers' work practices and identified three types of "fragmentation" that are motivated by tooling in the practice. Our study highlights several distinctions between traditional and contemporary information work settings and lays a foundation for future systems that aid next-generation information workers in managing work-related transitions.

We conclude by outlining this dissertation's contributions and future research directions.