Ancello Back to articles
How to Resolve Conflicting Research in a Shared Family Tree
6 July 2026 Ancello

How to Resolve Conflicting Research in a Shared Family Tree

Resolve a disputed family-tree fact by defining the claim, comparing source quality and identity evidence, and recording the reasoning.

Conflicting genealogy research is resolved by evidence, not by choosing the most popular tree or the most confident contributor. Freeze the disputed edit, define one precise claim, gather the sources behind every version, test whether they concern the same person, and document a conclusion that can remain uncertain.

Define the exact conflict

“We disagree about John Kelly” is too broad. Break it into claims:

  • Was John Kelly born in 1872 or 1875?
  • Were his parents Patrick Kelly and Mary Cain?
  • Is the man in the 1901 Douglas census the same man who married in Peel?
  • Did he die in 1938 or emigrate?

Resolve one claim at a time. Different dates may be a small reporting error; different parents may indicate two people have been merged.

Preserve both versions

Before editing, capture the current values, contributors, citations, notes, and attached records. Do not delete the minority view or detach sources merely to make the tree look tidy. If the platform supports change history, use it; otherwise create a dated conflict note.

Pause further merging and copying until identity is examined. Every new derivative tree makes an early error harder to trace.

Build an evidence table

List each source in its own row with:

  • source type and date created;
  • exact information relevant to the claim;
  • whether you saw an original, image, transcript, or index;
  • informant and likely knowledge, if known;
  • time between event and record;
  • identity details: relatives, occupation, address, age, religion, signature;
  • conflicts and limitations;
  • citation and file link.

Do not count copies as independent evidence. Ten online trees repeating one unsourced date are one tradition, not ten confirmations.

Test identity before comparing facts

Two records can each be accurate but concern different same-name people. Compare clusters rather than names alone:

  • spouse and children;
  • parents and siblings;
  • witnesses, neighbours, and associates;
  • occupation and employer;
  • precise address and movement;
  • age pattern across records;
  • signature or handwriting;
  • religious community;
  • property or military identifiers.

Create timelines for each candidate. Overlapping events in distant places can reveal a mistaken merge.

Evaluate the information, not only the document

An official certificate is not equally reliable for every field. A parent may report a child’s birth accurately; a neighbour reporting a death may guess the birthplace or parents. A census age can be approximate. A later obituary may combine memory and family tradition.

Ask who supplied each fact, how they knew it, why the record was created, and whether the information was recorded near the event.

Recheck the images

Return to the best available image or original. Read headings, marginal notes, adjacent entries, and page context. Indexes and automated hints are discovery tools, not substitutes for the record. Transcribe disputed text independently and compare readings.

For difficult handwriting, identify repeated letter forms elsewhere on the page. For a place or occupation, use period vocabulary rather than forcing a modern interpretation.

Search for evidence that can discriminate

Do not simply gather more records of the same weak kind. Ask what source would distinguish the hypotheses. A marriage record may name a father; a will may identify children; a military file may contain next of kin; land or directory records may separate addresses; DNA evidence may support a relationship when analysed with documented trees and consent.

Write a targeted research plan and log negative searches.

Reach a proportional conclusion

Possible outcomes include:

  • one version is supported and the other is demonstrably wrong;
  • one is more probable but a key record is missing;
  • both dates describe different events, such as birth and baptism;
  • two same-name people were merged;
  • the evidence is genuinely unresolved.

Use language that matches the result: “proved,” “probably,” “possibly,” or “not established.” Do not convert “most likely” into an unqualified tree fact.

Record the reasoning

Write a concise conclusion note:

  1. disputed claim;
  2. competing versions;
  3. sources examined;
  4. identity analysis;
  5. conflicts and how they were resolved;
  6. conclusion and confidence;
  7. remaining questions;
  8. date and contributors.

Link it to the person and the specific fact. Future relatives should not have to repeat the debate.

Keep collaboration respectful

Critique a claim, not the person who added it. Begin with shared goals: a trustworthy tree and preserved evidence. Invite the contributor to explain a private source or family tradition that is not yet visible. Agree that conclusions can change when new evidence appears.

A useful message is: “The 1901 record appears to describe a second John Kelly because both men are recorded in different parishes on the same night. I’ve linked the two images and timelines. Can we review whether the records were merged?”

Prevent repeat conflicts

  • Require citations for major relationship changes.
  • Keep research notes with unresolved facts.
  • Review merges and imported GEDCOM batches.
  • Use stable person identifiers.
  • Preserve change history.
  • Teach collaborators to distinguish hints, indexes, images, and conclusions.

The aim is not a tree without disagreement. It is a tree where disagreement is handled transparently and every conclusion remains connected to the evidence that earned it.

Sources and further reading