* Dealing with AGE anomalies

[Little.auk]

Following on from the Standard Names question, I was wondering how FH users deal with the situation where an individual gives different ages for events over time (e.g Marriages and censuses).

In particular which age should I record in the AGE field in the Facts Property Box? Should I use the age given in the source, or the true age at the time of the event?

I use AS for creating sources - and I use the Auto Text facility to create the "Text from Source" entry, which I then edit to give a true transcription of the source - warts and all.

I have a particularly good (or should it be bad!) example with my Paternal Grandmother - Her birth certificate, baptism record and first two census entry (1881 and 1891) all agree that her birth was in 1879 (June on the certificate).

She married in October 1896, and inflated her age by 2 years, to 19. (her husband was 10 years older than her).

Over the years, on censuses and the 1939 register, she gave various ages which were between 3 and 6 years older than her true age. It was not until her death certificate that her true age re-appeared!

[ColeValleyGirl]

I always put the age as given in the source, even if it's patently inaccurate. FH will display it with a (!) in the fact list to highlight the anomaly if you have a birth date.

[Gowermick]

Regardless of the age problem, I use the Birth date as given on GRO, Baptisme etc. (The latter taken with a pinch of salt, as baptisms aren't always carried out on babies). I narrow the date down as further evidence appears. So it may start as 'C 1825', then become '1827' then finally '4 May 1827'

When it comes to ages recorded on various events, I use the age that is given, even if is wrong. i.e. for a census, I am recording what appears on the census, not what I think it should be. I may occasionally add a note explaining why I think this is correct person, despite having an incorrect age.

Ages at death are notoriously unreliable, as the informant is only stating what they believe to be true, which may only be what deceased has told them, and who may have knocked a few years off.

I have found that the further an event is from the birth, the more unreliable the age given

[jbtapscott]

I also use AS for capture and always input the age as shown on the source document. On some occasions I add a Note to the fact where I think a bit more information may be helpful.

On my website (built using GedSite) I don't include Age in the sentences for facts (the info can be seen in the "Text from Source" field on the Source record if somebody is particularly interested), but have included a comment in my "About" section to the effect that any Ages that are shown are taken from the original source document, etc.

[Little.auk]

Thanks all,

I record the 'Source given' ages in all my sources, plus any name anomalies (like ROLLIN transcribed as BOLLIN), but was not sure the best way to go for the 'Fact Box' age, but your consensus is clearly to use the age from source.

Regarding the unreliability of ages on death certificates - I think this has become less of a problem for more recent years. My grandmother died in 1955 and her death certificate gives the first correct age for her since her marriage.

[davidf]

I always put the age as given in the source, even if it's patently inaccurate. FH will display it with a (!) in the fact list to highlight the anomaly if you have a birth date.

But you can only be certain that it is an anomalous date if you are certain (and precise) about the birth date. I'm not sure about how the anomaly "! flag" is determined given that for many events we do not know if age is "age last birth day" - which I guess is what most of us give - or age rounded to the nearest year.

I get caught by the FH warning if I enter a baptism for 14 Feb 1881 and then a ~DoB from GRO as Q1 1881. That presumably is because Q1 1881 is interpreted as 1/1/1881 to 31/3/1881 - which just means a possible error - despite the nag! Entering the two events in the opposite order (birth date range, followed by precise baptismal date) does not appear to give the nag - because FH sees no inconsistency - despite the possibility of error.

But if a DoB is entered as Qn YYYY (or presumably any form of date range) that will potentially throw the date anomaly "! flag" for every event.

Regarding the unreliability of ages on death certificates - I think this has become less of a problem for more recent years. My grandmother died in 1955 and her death certificate gives the first correct age for her since her marriage.

Isn't it more to do with how well informed the informant is? If a person went into a care home and on entry incorrectly gave their date of birth, a care home manager as informant can only go with what they were given.

My late mother's birth details were never checked when she went into a care home - and I was never asked to confirm the details. Likewise when she died, I attended the registrar with both her birth certificates and her marriage certificate, plus my own passport ready to prove who I was. The registrar was only interested in whether I was the bearer of the doctor's letter - she took everything else on trust - but she chose the occupation of "housewife" from the range of options that I presented.

[ColeValleyGirl]

I always put the age as given in the source, even if it's patently inaccurate. FH will display it with a (!) in the fact list to highlight the anomaly if you have a birth date.

But you can only be certain that it is an anomalous date if you are certain (and precise) about the birth date. I'm not sure about how the anomaly "! flag" is determined given that for many events we do not know if age is "age last birth day" - which I guess is what most of us give - or age rounded to the nearest year.

It's an anomaly in that it doesn't fit in with the putative birth date, nor possibly with the age given at other events. it doesn't help you to know which is more accurate, just that there's a discrepancy.

[dewilkinson]

I always use the age as given in the source, but if it is quite a bit out I will add a note in the text so I know I have already questioned it. I think it is important to record as given.

[Martin Tolley]

My great grandmother had six different dates of birth, all recorded on "official" documents of one sort or another. There are different days, different years and different months. I don't think she ever knew what the real one was. My great grandfather told me that she chose to celebrate her birthday on 25 June as she thought it was nice to have a party 6 months after Christmas, as it broke up the year.

[Gowermick]

I get caught by the FH warning if I enter a baptism for 14 Feb 1881 and then a ~DoB from GRO as Q1 1881.

Just do what I do, and change and correct the birthdate as more evidence appears. 1881 Q1, is quite broad, so correct it , when you find the baptism, to bef 14 Feb1881, which is more specific.
Don’t forget, people had 42 days approx to register a birth, so births registered in 1881 Q1 could have been born anywhere from mid Nov 1880 to end of March 1881!

[davidf]

Just do what I do, and change and correct the birthdate as more evidence appears. 1881 Q1, is quite broad, so correct it , when you find the baptism, to bef 14 Feb1881, which is more specific.
Don’t forget, people had 42 days approx to register a birth, so births registered in 1881 Q1 could have been born anywhere from mid Nov 1880 to end of March 1881!

I certainly gather the data - and a date of birth in my projects can have multiple sources quoting or (for the likes of census and marriage records) implying different dates - and I put the "most likely" value in the actual field with its source in Source[1] position.

However to be picky bef 14 Feb1881 (the baptism date), is not necessarily more specific than 1881 Q1. The Quarter date combined with the baptism date implies a probable birth date between say 24 December 1880 and 14 February 1881 - taking Christmas Eve as the probable last date of registration for Q4 births. There is as you point out the 42 day limit which would push the possible range out to 19 November 1880 to 14 February 1881 - which would exclude adult baptism. The choice of bef 14 Feb 1881 could include a birth well before - as for an adult baptism - or batch baptisms of "broods".

I have the same qualms about using a bef [Burial Date] rather than a GRO Quarter date.

Possible we need a "close before" date facility? But then how do we define close?

I can live with Quarter dates - but then I do not use FH to "auto produce" output with the implied precision of a number on screen or paper; I use it as a repository that I then interpret when writing Family History.

[Gowermick]

David,
You can discount adult baptism in this case, as you have a Gro date of birth of Q1 1881 and so the christening in Feb 1881 indicates an infant christening.

[tatewise]

You can use a Range Date such as btw Dec 1880 and 14 Feb 1881 which well defines the end date while leaving the start date slightly open although indicating your required closeness.

The same strategy can be used where several sources suggest a number of dates.

[KFN]

For me it depends heavily on how sure I am of the actual age of the individual at the time of the dated event.

If I know the age at the event with a high degree of accuracy then the age given in the source and the text source goes in the TEXT field and my accurate age goes in the AGE field.

If I don’t know the age at the date of the event then the source age goes in the AGE field as well as in the TEXT field (in case I need to adjust the data later).

I see this type of field as a “needs to be accurate” field, but use fields like TEXT and NOTE as a place to put reported information and notes to explain my actions.