Interrupts

TL;DR: Random thoughts on interruptions and how to handle them. Probably not that useful or interesting to others yet.

I am interested in thinking more formally about interruptions in my daily life and how to handle them. I’m reading a bit about this.

It seems to be well-known that interrupts are costly to the individual. “Flow state” is generally recognized as a good thing. Developers complain about too many meetings. Books like “Deep Work” - Cal Newport and essays like Maker’s Schedule, Manager’s Schedule by Paul Graham lament interruptions.

However, I am finding it hard to square the individual-level costs of interrupts with the group-level benefits. I cannot recall the exact quote, but David Allen and Edward Lamont’s book “Team” says somewhere that responsiveness is correlated with some job success metrics. I fail to recall which ones. But this seems reasonable — people sometimes need information from me, and sometimes their work is blocked until I can provide it. There are lots of ways to pre-empt this problem, and I’m working on implementing some of them. At the same time, I don’t think I can avoid interrupts completely. I’m curious if it would help to have a better understanding about interrupts to feel comfortable handling them.

In “Leader responsiveness, equity sensitivity, and employee attitudes and behavior” by Shore, Sy, and Strauss, the authors hypothesize that leader responsiveness correlates with employee job performance. This was cherry-picked as an example of some prior work suggesting a correlation between responsiveness and positive group-level outcomes.

Computer science has a very developed theory for how interrupts are handled by the operating system and computing hardware. It seems there is less work on how humans handle interruption.

This seems like something amenable to computational research.

Interruption science as a research field: Towards a taxonomy of interruptions as a foundation for the field Fabian J. Stangl and René Riedl 2023 Frontiers in Psychology

Authors seem to be business school professors.

I like this list of defining factors for what constitutes an interruption, from (Puranik, et. al. 2020):

1. Suspension of an ongoing task’s execution
2. Unexpectedness of its occurrence
3. Presence of an interrupting task
4. Intention to resume the interrupted task
5. Interruption source is internal or external

From “The Morning Inbox Problem: Email Reply Priorities and Organizational Timing Norms** by Heejung Byun and David A. Kirsch:

> To which emails do employees first respond? This paper explores the near-universal element of being an employee responsible for attending to an organizational email account. Using a large email corpus of an organization across five years, we explore various response heuristics that might be guiding how organizational members make reply priority decisions. Study 1 simulates what we call the “morning inbox problem”—deciding whom to reply to when emails from multiple respondents have been received. Our analysis reveals that **organizational members use the following heuristics in reply priority decisions: (a) order of arrival, (b) contents that can be inferred from the subject line rather than from the email body, (c) effort required to reply, and (d) expectation the sender has with respect to time-to-response of the email.** These patterns of reply priority reveal a new construct—organizational timing norms, which are norms that govern the temporal dimension of intraorganizational interactions. Study 2 explores the organizational implications of timing norm violations. We show that **organizational members sending out-of-sync replies constitute a “bottleneck” in organizational communication** and is a predictor of employee turnover. We discuss the implications of our discoveries and underscore the potential value of studying timing norms in organizational routines.

An interesting title from Laura M. Giurge and Vanessa K. Bohns. This tracks with my personal feeling.

“You don’t need to answer right away! Receivers overestimate how quickly senders expect responses to non-urgent work emails”

I am realizing that I have put all the work on myself in inferring priority/urgency/impact for incoming messages. Some of this could probably be done by the sender if we can agree on a convention. I’ll keep this in mind as something to talk about with frequent collaborators.