Attention residue: what follows you into the next task
The finding
Sophie Leroy's 2009 paper describes attention residue as thoughts about a prior task that persist while working on another. Across two experiments, difficulty disengaging from unfinished work affected subsequent performance. Simply finishing was not enough: time pressure while finishing helped participants disengage. This is narrower than saying every switch damages work in the same way.
The handoff study is different
In 2018, Leroy and Theresa Glomb examined interruptions when people anticipated resuming unfinished work under time pressure. Across four studies, that expectation was associated with attention residue and poorer performance on the interrupting task. A brief ready-to-resume plan mitigated the effect. It belongs to this later paper, not an invented third experiment in 2009.
A practice to try
Leave a short note before changing projects: the next action, the obstacle and where to restart. This exact format is an editorial suggestion, not the experimental protocol or a measured software benefit. Try it on work you genuinely need to resume. Neither cited paper establishes universal minutes lost per app switch or validates a particular workspace app.
What a transition note could contain
Consider a hypothetical writer leaving an outline to answer a necessary message. A useful note might say: "Next: check the source for the second paragraph. Open question: does the evidence describe a task or a whole working sphere? Resume at section two." This is an example, not testimony from Roberto or a study participant. It preserves a decision that would otherwise have to be reconstructed.
The format need not be elaborate. Avoid turning the handoff into another project-management chore. For a proposed personal test, compare whether you can state the next action when you return, with and without the note. Record your own observation rather than claiming the literature already measured that particular workflow. A saved set of windows might help you find material, but the cognitive and software questions remain separate.
Sources
Leroy 2009 publisher abstract, two experiments
https://www.sciencedirect.com/science/article/abs/pii/S0749597809000399
Leroy & Glomb 2018 publisher abstract, four studies
https://pubsonline.informs.org/doi/abs/10.1287/orsc.2017.1184