# Evidence and sources

Research checked 10 September 2026. This is a focused research synthesis, not a systematic review conducted by Polar Bear. The eight workflows are original design applications. No study evaluated these prompts or established their effectiveness for agencies.

## What the evidence supports—and does not

Errors create information, but learning depends on attention, analysis, resources, and implementation. Evidence supports studying conditions and testing changes rather than treating an adverse outcome as proof of personal incompetence. Training experiments concern protected practice, not deliberately allowing harmful real-world errors. Research on psychological safety is largely associational; respectful responses do not guarantee reporting or remove accountability. Healthcare incident studies offer useful mechanisms, but their outcomes cannot be carried directly into professional services.

Each entry provides its access level, evidence type, design implication, and principal limit. Abstract-only entries support only broad findings reported in their abstracts. Full-text sections means relevant portions were retrieved, not that every page was read.

| ID | Academic source | Evidence and access | Use and limitation |
|---|---|---|---|
| E1 | Dahlin, K. B., Chuang, Y.-T., & Roulet, T. J. (2018). Opportunity, Motivation, and Ability to Learn from Failures and Errors. *Academy of Management Annals*, 12, 252–277. [DOI](https://doi.org/10.5465/annals.2016.0049) | Integrative review; abstract and author manuscript opening. | Reconstruction and causal mapping: distinguish process quality from outcome. Framework synthesis, not a tested incident protocol. |
| E2 | van Dyck, C., Frese, M., Baer, M., & Sonnentag, S. (2005). Organizational Error Management Culture and Its Impact on Performance: A Two-Study Replication. *Journal of Applied Psychology*, 90, 1228–1240. [DOI](https://doi.org/10.1037/0021-9010.90.6.1228) | Two organizational field studies; university-hosted abstract. | Receiving reports and repair: detection, communication, correction matter. Associations in Dutch and German organizations do not establish causal effects. |
| E3 | Keith, N., & Frese, M. (2008). Effectiveness of Error Management Training: A Meta-Analysis. *Journal of Applied Psychology*, 93, 59–69. [DOI](https://doi.org/10.1037/0021-9010.93.1.59) | Meta-analysis of 24 studies; university abstract and manuscript opening. | Prevention practice: protected exploration can support transfer. Effects vary; not permission to experiment on clients. |
| E4 | Martin-Delgado, J., Martínez-García, A., Aranaz, J. M., Valencia-Martín, J. L., & Mira, J. J. (2020). How Much of Root Cause Analysis Translates into Improved Patient Safety: A Systematic Review. *Medical Principles and Practice*. [DOI](https://doi.org/10.1159/000508677) | Systematic review, 21 studies; full-text methods/results/discussion sections retrieved from [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC7768139/). | Prevention tests and closure: recommendations and implementation are recurring weaknesses; only limited studies demonstrated safety improvement. Heterogeneous hospital research, much observational; do not infer that analysis alone prevents recurrence. |
| E5 | Peerally, M. F., Carr, S., Waring, J., & Dixon-Woods, M. (2017; online 2016). The Problem with Root Cause Analysis. *BMJ Quality & Safety*, 26, 417–422. [DOI](https://doi.org/10.1136/bmjqs-2016-005511) | Scholarly critique; abstract/opening retrieved. | Multi-factor mapping and implementation checks. Conceptual critique, not comparative intervention proof. |
| E6 | Reason, J. (2000). Human Error: Models and Management. *BMJ*, 320, 768–770. [DOI](https://doi.org/10.1136/bmj.320.7237.768) | Conceptual article; opening and summary verified. | Examine working conditions and barriers alongside actions. The model is a way to inquire, not a mechanism for assigning blame or proving a unique cause. |
| E7 | Frese, M., & Keith, N. (2015). Action Errors, Error Management, and Learning in Organizations. *Annual Review of Psychology*, 66, 661–687. [DOI](https://doi.org/10.1146/annurev-psych-010814-015205) | Review; indexed metadata and abstract excerpt. | Combine prevention with detection and recovery. Broad review does not validate this pack’s sequence. |
| E8 | Breines, J. G., & Chen, S. (2012). Self-Compassion Increases Self-Improvement Motivation. *Personality and Social Psychology Bulletin*, 38, 1133–1143. [DOI](https://doi.org/10.1177/0146167212445599) | Four experiments; [PubMed abstract](https://pubmed.ncbi.nlm.nih.gov/22645164/). | Optional fair self-description alongside repair. Short-term experimental tasks and motivation do not prove enduring workplace recovery. |
| E9 | Frazier, M. L., Fainshmidt, S., Klinger, R. L., Pezeshkan, A., & Vracheva, V. (2017). Psychological Safety: A Meta-Analytic Review and Extension. *Personnel Psychology*, 70, 113–165. [DOI](https://doi.org/10.1111/peps.12183) | Meta-analysis; university repository abstract and manuscript opening. | Respond respectfully to reports and preserve appropriate routes. Much underlying evidence is correlational; no promise of safety or confidentiality. |
| E10 | Tucker, A. L., & Edmondson, A. C. (2003). Why Hospitals Don’t Learn from Failures: Organizational and Psychological Dynamics That Inhibit System Change. *California Management Review*, 45, 55–72. [DOI](https://doi.org/10.2307/41166165) | Qualitative field research in nine hospitals; publisher abstract and article opening. | Close reporting loops and transfer process changes instead of relying solely on local workarounds. Context-specific observational findings. |
| E11 | Eskreis-Winkler, L., & Fishbach, A. (2019). Not Learning From Failure—the Greatest Failure of All. *Psychological Science*, 30. [DOI](https://doi.org/10.1177/0956797619881133) | Five experiments; journal-society abstract. | Do not presume failure automatically teaches. A key paradigm is disputed (E12); do not present ego threat as a settled universal explanation. |
| E12 | Spiller, S. A. (2025). Commentary on Eskreis-Winkler and Fishbach (2019): A Tendency to Answer Consistently Can Generate Apparent Failures to Learn From Failure. *Psychological Science*. [DOI](https://doi.org/10.1177/09567976251333666) | Methodological critique plus experiment; full-text results/discussion sections retrieved from [PMC](https://pmc.ncbi.nlm.nih.gov/articles/PMC13428910/). | Qualify E11: response consistency can mimic learning differences in the Script Task. Does not settle every failure-learning finding or imply failure never affects learning. |
| E13 | Loh, V., Andrews, S., Hesketh, B., & Griffin, B. (2013). The Moderating Effect of Individual Differences in Error-Management Training: Who Learns From Mistakes? *Human Factors*. [DOI](https://doi.org/10.1177/0018720812451856) | Randomized simulated task experiment; indexed metadata and abstract. | Use scaffolding and test transfer. Error encouragement exceeded avoidance but not the control in the abstract; it is not uniformly superior. |

## Original design choices

Containment cards, accountability wording, timeline columns, action ownership, exposure denominators, rollback fields, and closure labels are practical synthesis choices. Research motivates checking implementation and uncertainty; it does not establish these specific templates or time limits. An apology draft is practical communication design, not a research-backed guarantee of trust repair. A two-minute reflection is optional convenience, not a tested dose.

## Research method and boundaries

Firecrawl’s paper index was searched with separate organizational-learning and error-training queries; related-paper expansion from the training meta-analysis identified relevant boundary-condition studies. In-body retrieval was attempted for load-bearing papers, but the index returned no text. Ordinary web research retrieved primary publisher/university material and full-text sections where available. We prioritized reviews, organizational field research, relevant experiments, and counterevidence. We excluded off-topic computational learning papers and did not transfer clinical procedures into management instructions. No paywalled full text was inferred from abstracts.