The Field Guide › Paper
An integrative review of the organisational-error literature (some sixty-five studies across twenty-five years) that reads error research through three lenses and, in doing so, becomes the corpus's anchor for the error-management tradition running alongside the Edmondson learning-from-failure lineage. It first fixes the object: errors are unintended, potentially avoidable deviations from goals or standards, distinct from violations (intentional rule-breaking), failures (bad outcomes, which errors may or may not cause), and risks (environmental hazards). Against that definition it sets the field's central fault line, between an error-prevention paradigm that treats errors as wholly negative and works through routine, standardisation and control, and an error-management paradigm (Frese and Keith) that treats them as feedback and as openings for learning and innovation, and works through adaptation, flexibility and quick correction. The three lenses organise the synthesis. The level-of-analysis lens sorts errors into individual, team, and organisational or system errors, and shows their antecedents interacting across levels. The temporal lens follows an error over time, through before, during and after phases, and takes in latent errors (Reason) that lie dormant until other conditions trigger them, error-amplifying and error-correcting feedback loops, and the timing of responses, where both too-slow and too-fast reactions can breed further error. The priority lens examines what happens when error elimination competes with other goals, and it is here the review is sharpest: strategies that suppress error at one level or phase routinely raise it at another, and the positive effects of a given error practice tend to turn negative when pushed too far. The running theme is non-monotonicity and tension rather than a tidy set of best practices; one striking example is the finding (Katz-Navon and colleagues) that a high priority of safety combined with a high active-learning climate produces more errors, not fewer, because the two pull against each other. Psychological safety features throughout, but as one team-level antecedent among several: the review draws on Edmondson (1996) linking psychological safety to open reporting of error, on Leroy and colleagues finding that a safety priority reduces error most where psychological safety is high, and on Nembhard and Edmondson on the tendency to cover errors up, reading all of these through the error-management frame rather than the psychological-safety one. Its place in this corpus is as a review-level bridge: it connects the psychological-safety and learning-from-failure work to the wider error-management, high-reliability and normal-accident literatures (Goodman et al., Perrow, Weick and Sutcliffe), and supplies the multilevel, temporal and priority vocabulary that the corpus's error material had lacked a single home for.