BPMN Model Validation Survey

Wikipedia Dispute Resolution — Capstone Team Expert Review

This survey is part of the Wikipedia Dispute Resolution research project. The capstone team has produced a series of BPMN process models of Wikipedia's dispute resolution system, generated programmatically from case metadata (parties, votes, dates, principles, findings, and remedies).

We need your expert feedback to validate these models against how the process actually works in practice. Please review the BPMN diagrams on the main project page before answering, and respond to all questions. Your answers are saved anonymously and can only be downloaded by administrators as a CSV.

**How to use this survey:** Open the [[wikipedia_dispute_resolution|Wikipedia Dispute Resolution page]] in a separate tab so you can refer to the BPMN diagrams while answering. The overview diagram shows the full escalation flow; the detailed models for each stage are linked under the "Detailed Process Models" section.

—-

Q1 — Overall Completeness: Does the overall diagram capture the full process from start to finish?

Yes, it appears complete Mostly complete, with minor gaps Partially complete — significant steps are missing No, it is substantially incomplete

Q1a — Are there any steps that happen in practice but are missing from this diagram? (Leave blank if none)

Q1b — Are there any decision points where the process branches that are not shown here? (Leave blank if none)

Q2 — Flow Order: Does the left-to-right order of steps match how cases actually progress?

Yes, the sequence is accurate Mostly accurate, with minor ordering issues Some steps are in the wrong order The overall ordering does not reflect reality

Q2a — Are there steps shown in sequence that can actually happen concurrently or in either order? If so, which ones?

Q2b — Are there points where a case can loop back to an earlier step (e.g., a decision gets revisited, a phase reopens) that the diagram does not show? If so, where?

Q3 — Decision Gateways and End Events: Do the branching conditions and outcome labels on the gateways (diamonds) and end events (circles) match reality?

Yes, they accurately reflect real conditions and outcomes Mostly accurate, with minor labelling issues Some labels or conditions are inaccurate or missing The gateways and outcomes are significantly wrong

Q3a — Are the outcome categories (e.g., Remedies Imposed, Declined, Resolved, No Consensus) the right buckets, or would you group them differently? Please describe any changes you would make.

Q3b — Are there outcomes that actually occur but are not represented as an end state in the diagram? If so, what are they?

Q3c — For each gateway, are the labels on the outgoing arrows (e.g., Accepted / Declined) the right conditions, or are there other paths a case can take from that point? Please note any incorrect or missing arrow labels.

Q4 — Swimlane Attribution: Is each task placed in the correct lane (attributed to the right actor)?

Yes, all tasks are correctly attributed Mostly correct, with a few misattributions Several tasks are in the wrong lane The lane structure does not reflect how roles work in practice

Q4a — Are there specific tasks placed in the wrong lane? If so, which tasks should be moved, and to which actor?

Q4b — Are there roles or actors that do not have their own lane but should (e.g., bots, specific committee roles, uninvolved editors)? If so, which roles are missing?

Q5 — Data Extraction Quality: The model was generated from programmatically extracted metadata (parties, votes, dates, principles, findings, remedies). Overall, how well does this extracted data represent the actual process?

Well — the extracted data gives a reliable picture Partially — some important dimensions are missing or distorted Poorly — the extracted data is substantially misleading Unable to assess

Q5a — What key data points are we failing to capture that would be important for a complete picture of the process?

Q5b — Are any of the data points we did extract misleading or unreliable as parsed (e.g., vote counts that do not mean what we think they mean)? If so, please describe.

About you — What best describes your relationship to Wikipedia's dispute resolution processes? (Optional — helps us interpret your feedback)

Active Wikipedia editor with DR experience ArbCom member (current or former) DRN or Mediation Committee volunteer Administrator with DR involvement Researcher / outside observer Prefer not to say

Any other comments or observations about the BPMN models not covered above? (Optional)

You need to be logged in to answer this questionnaire

About This Survey

This survey is part of the Wikipedia Dispute Resolution capstone project. The BPMN models it references were generated programmatically and are documented on the main project page alongside process component descriptions, key actors, and policy cross-references.

Useful links for reviewers:

Results will be exported as CSV for analysis by the capstone team. Thank you for your time.

You could leave a comment if you were logged in.