Potential improvement to DB error 210: Father was dead before birth

Question

In the Database Error Project, error 210 Father was dead before birth is defined as "This person's father died before birth so probably birth date or father's death date is wrong."  I know of four people in my direct family tree who were born after their father died in military service or on car accidents.  I also have one potential problem whee the birth year of the child is estimated based on the census record and the father's death date is based in the tombstone and cemetery record.  This requires more research int the child's birth date, which may be either before or after the father's death - I need the month to know for sure.

To reduce the number of false errors (and potential false corrections), could we consider changing the edit check to flag only those who were born more than 1 year after their father's death?

Answer 1

Are you aware it is already set at 10 months?

Comments

I was not aware of that.  Thanks for letting me know.

Answer 2

I have come across the same "error" several times now. Many times the date of birth is not known but an estimate of the date of baptism. So frequently children would be baptised a week, month[s], years after or even at marriage (for many reasons - living on isolated farms, illness, availability of preachers and or settings, especially for those who were on "trek"). So there are many many profiles of the date of baptism with the "before" option selected. I know there was a G2G feed on this whole before / after approach of the error database. Now I recently went through my own personal errors list after "hiding" some of these specific errors (parents having died before a childs birth) more than a month before. If I saw correctly, after "hiding" such an error the option for "false" error has disappeared. So now I cannot (after having had time to take a good look at the data) "correct" the "error" as a "false" error. This is something that I would like to have clarified - is this true?

Another point I'd like to make - recently we had a spout (in our project COGH) of a correction officer's well-intentioned but misguided manipulating of data to accomodate the error database. Without comments and or with comments. For example:

* Dies before marriage
* Child [name] born before marriage
* All other children born after death
* [...] only child not born in [county name]
* No baptism record of [...] can be found in the [wrong archives]

'''Please research further, or at least record these current contradictions in the profile, if only to prevent a "Correction officer" to make unsourced corrections to the profiles!'''

This specific person (not the speciific one that was bulldozing through [also validated] data) was kind enough to leave and extended message, but in his comments it is evident that he had not read the extensive analysis (footnotes / research) of some anomalies regarding the baptism record that could not be found, in the biography, even though there were real issues. A simple date of death error corrected solved some of the errors. And there was a simple mistake regarding the marriage. It's great to have an error database to help clear up these anomalies, but in no single instance should a correction officer make "unsourced" corrections to profiles. See these feeds Does everyone READ the comments ... and Guidelines for working Database errors and Can someone tell what a correction officer does?. Thanks!

Comments

There is an option, before you run the report, to un-hide hidden or false errors, just so that you can recheck them

http://www.sdms.si:92/wikitree/ShowErrors.htm

---

For error 210 I removed possibility to mark error as false error, since there is no reason this option is needed. Date of baptism isn't the day of birth. So if one were baptised at age 10 and his father died 5 years before that, you can make a better guess of birth date according to father's death + 1 year and before option. Otherwise other errors can come from such guess. (He might have a child at age of 5,...).

---

Thanks, I did not know that. But even after the period had expired and they had become "visible" again, the option to set it as a "false" error had disappeared ... or perhaps I was not doing the right thing ...

---

Thanks for your honest answer Ales, I appreciate it.

But then we have a huge problem. I'm horrified frankly. Thousands of profiles have the dates of birth aprroximated to the date of baptism, and in our project alone (+ 10000 profiles) it is not viable and not the best option to validate this way [edit: the way you describe, though with time every profile will have more or less when a exact date of birth is not availible, be making use of this method], let alone explain [how - because even when explaining the "error" will still remain] just to accomodate the error database ...

---

I think that source comment in bio Birthdate estimated based on father's death date. is all that is needed in addition to such a guess. There will never be possible to make a better one.

I did see a lot of estimated dates due to lack of sources with even less accuracy. 

I am actually missing date range on wikitree, since such guesses could be limited on both ends. We have daterange for a year 1954 or for a month June 1954, but that is not very useful for such cases.

---

I think what is being said here is that we will NEVER get rid of all the errors.   Some of them will stay because research is still pending. And as long as I do not get a comment every week that says the same thing...RESEARCH needed, I am OK with that.