How to Find Conflicting Dates in a GEDCOM File

Hard impossibilities — fix these first

• Birth year after death year for the same person. Almost always a data-entry error — a transposed digit, or a date typed into the wrong field.
• Marriage recorded after a spouse’s death date. Check whether the marriage date or the death date was mistyped, or whether the marriage is actually linked to the wrong person.
• A child born when a parent would have been younger than 10. Biologically implausible in any era — usually means the child is linked to the wrong parent, or a generation was skipped when the tree was built.

Implausible, not impossible — worth a second look

• A child born when the father would have been older than 80, or the mother older than 60. Not impossible, but unusual enough to double-check the birth date, the parent identity, or whether two same-named individuals were merged into one person.
• Events on the same person showing cross-continental travel within 3 days of each other. Before the 20th century, transatlantic travel took weeks. Two same-named people in different countries in the same decade almost always means two different individuals were accidentally merged into one record.
• Five or more dated events spanning three or more continents with no residence record connecting them. The same accidental-merge pattern, just more spread out.

How to track these down in your own software

1. Sort or filter individuals by birth/death date to catch any death-before-birth cases first — these are unambiguous and the highest priority.
2. For flagged parent-child pairs, check the parent’s age at the child’s birth. Most genealogy software can show this directly, or you can subtract birth years.
3. For anyone with events in widely different places close together in time, check whether you’re actually looking at one person or two people who share a name and got combined.
4. Where a date really is correct but looks unusual — a modern date sitting next to historical records, for instance — add a note explaining why, so it doesn’t get re-flagged as an error later.

What this kind of check can’t tell you on its own: which of two conflicting sources has the transcription error, whether an unusual date is a mistake or an intentional historical anomaly, or whether a cross-continental pattern is one person or two merged individuals — those all need your own review of the underlying records.

Doing this automatically

KinSift’s Evidence Analysis check scans your exported GEDCOM for exactly these patterns — chronological impossibilities, implausible parental ages, and the cross-continental travel signature — and lists each one against the specific people involved, for free.