Trigger Analysis
understanding broken tasks


Alan Dix

Devina Ramduny-Ellis

Julie Wilkinson


In The Handbook of Task Analysis for Human-Computer Interaction.
Eds. Dan Diaper & Neville Stanton. Lawrence Erlbaum Associates, 2004.

Full reference:
A. Dix, D. Ramduny-Ellis, J. Wilkinson (2004). Trigger Analysis - understanding broken tasks. Chapter 19 in The Handbook of Task Analysis for Human-Computer Interaction. D. Diaper & N. Stanton (eds.). Lawrence Erlbaum Associates, 2004. pp. 381-400
Download draft chapter (PDF, 113K)
Alan's pages on ecology of interaction and time

Why do things happen when they happen? Trigger analysis exposes the triggers that prompt sub-tasks to occur in the right order at the right time, and assess whether tasks are robust to interruptions, delays and shared responsibility. Trigger analysis starts with task decomposition via any suitable method, proceeding to uncover what trigger prompts each sub-task. The obvious answer: "because the previous sub-task is complete" is often the precondition, not the actual trigger. Previous analysis has uncovered certain primary triggers and potential failure points. The complete analysis produces an ecologically richer picture, with tasks interacting with and prompted by their environment.

keywords: sub-tasks, triggers, placeholders, artefacts, shared tasks, environmental cues, task timing, interruptions, delays


Many task analysis and workflow techniques decompose the overall task into smaller subtasks, processes or activities. The order of these subtasks is then typically specified or described (e.g. plans in Hierarchical Task Analysis (HTA), links between processes in a workflow, temporal connectives in CTT). For short- term tasks, performed by one person, without interruption, this may be the end of the story:

to photocopy a document:
(a) open copier lid,
(b) put original on glass,
(c) close lid,
(d) select number of copies,
(e) press copy button,
(f) when copying complete remove copies,
(g) remove original

But what happens if:

Trigger analysis deals with exactly these issues. Why do things happen when they happen and do they happen at all? By exposing the triggers that prompt activities and sub-tasks to occur in the right order at the right time, trigger analysis allows us to decide whether tasks are robust to interruptions, delays and shared responsibility (even across organisational boundaries).

Trigger analysis starts with a task decomposition obtained by any suitable method (and can therefore be used in combination with many TA and workflow methods). It then proceeds to uncover what trigger causes each subtask to occur. The initial answer is "because the previous sub-task is complete", but this is often merely a precondition, not the actual trigger.

Previous empirical and theoretical analysis has uncovered a small set of primary triggers including the simple "previous task complete", timed events – "every hour I check the mail", and environmental cues – "the document is in the in-tray". For each class of trigger there are a set of subsequent questions, for example "what happens if you are interrupted between tasks?", "how do you know when it is the right time?", "are there several tasks with the same environmental cues?".

Triggers are what make activities happen when they do. A closely related issue is: knowing where in task sequence you are. Often, environmental cues act in both roles: triggers saying that something should happen and placeholder saying what should happen. Typically, the complete analysis produces a highly ecologically rich picture and rather than cognitively-driven tasks acting on the environment, we see tasks interacting with and prompted by their environment. Note that trigger analysis is not an alternative task analysis technique or notation but instead an additional concern that should be grafted to existing analysis methods.

This chapter examines the trigger analysis method in depth, with particular reference to its grounding in empirical work. The starting point is a brief discussion of the theoretical underpinning for the fieldwork undertaken. We then suggest some reasons why prolonged interactions tend to break down in organisational contexts, culminating in an explanation of the five trigger types that have emerged during our investigations. A second, related pattern – the 4Rs framework – is also presented in detail, as we suggest the 4Rs form an important and fundamental unit of work. Finally, we demonstrate a comprehensive application of the 4Rs and the benefits they provided for the analysis of work in a systems development project.

references and links

Alan Dix 2/7/2002