I think most of us would admit that there are likely errors in our family trees — especially when those trees are large.
Because I track many of my FAN Club members in my database, my tree is not small. It contains multiple collateral lines and quite a few “floating” branches. Some of those individuals have not been fully researched. In other cases, I may have relied on only one or two sources. That reality alone increases the possibility of error.
However, I do not expect to find errors in the information for my direct ancestors through my second great-grandparents.
So imagine my surprise.
As I was preparing to create AI-generated stories for my 2nd great-grandfather, Washington Marion Crawford, I decided to do a bit of fresh research to see if anything new had surfaced. When I used Ancestry’s search feature, I was not expecting to see references to Iowa sources connected to him.
After adjusting the search filters, I discovered that he was supposedly living in Van Buren County, Iowa.
That immediately raised a red flag. To my knowledge, this family never lived in Iowa. They moved from Indiana to Kansas. There was no Iowa stop along the way.
So I clicked into the 1880 residence fact attached to him in Ancestry to identify the source. The source was — correctly — the 1880 census for Indiana.
Which led to the obvious question:
How did Iowa get attached to an Indiana census record?
Because I use RootsMagic as my master database and generate my Ancestry tree from it, I suspected the error originated there. The most logical explanation is that I inadvertently merged two similarly named places — likely a Washington Township in Indiana with Washington Township in Van Buren County, Iowa.
If that happened once, it could have happened multiple times.
Time to investigate.
Using RootsMagic’s rule-based groups feature, I created a group containing anyone with an event whose place name included:
- Washington Township
- Van Buren County
- Iowa
The result? Twenty-three individuals.
Thanks to my color-coding system, I could quickly distinguish my Crawford line from my Crandall line — the Crandalls actually did live in Van Buren County, Iowa. That visual separation made it easy to focus only on the five individuals connected to my Crawford research.
For each of those five:
- I checked the original source.
- Verified that the record was indeed from Indiana.
- Corrected the place name in RootsMagic.
Once the corrections were made, I used TreeShare to push the updated information back to Ancestry.
Problem fixed — and likely prevented from spreading further.
This experience reminded me why Thomas MacEntee’s “Do-Over” concept is not a one-and-done event. A do-over is a mindset. It’s a commitment to continual review, re-evaluation, and refinement.
Even well-researched lines deserve periodic scrutiny.
And I’m grateful that RootsMagic’s rule-based groups made it possible to locate these incorrect place names quickly. Without that tool, I might have corrected Washington Marion Crawford — and never discovered the other four errors quietly hiding in my database.
Sometimes the real value of our software isn’t data entry.
It’s quality control.