A procedural question

[J. Hugh Sullivan]

I am posting this to soc.genealogy.methods also because it's the
proper place. But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.


I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

Discuss please...

Hugh

[singhals]

J. Hugh Sullivan wrote:

I am posting this to soc.genealogy.methods also because it's the
proper place. But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.


I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

Discuss please...

Hugh

I've had the same problem in PAF and FTM. My work-around is, don't run
the problem report. (g)

Seriously, if date fields would take a less-than or greater-than or the
approximately-equal symbols, it would help. At least, the formula would
function with them. I don't myself see why the programmer
couldn't/didn't/won't "interpret" bef to less-than and aft to
greater-than and be done with it, but it doesn't seem to have occurred
to them.

Cheryl

[Robert Heiling]

"J. Hugh Sullivan" wrote:

I am posting this to soc.genealogy.methods also because it's the
proper place.

If you double-check the respective charters, you'll find that computing is
the right place. I note that methods gets a lot of posts that don't belong
there.

But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.

I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755".

Just as an aside, "Bef 1755" may not be correct as he could have been born
in 1755 and still have been 45+. I'd use Abt 1755.

From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Wasn't he?<g> They could more accurately haver said "possibly born after
marriage" as Bef 1755 could mean sometime before the Abt 1724.

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

Well *possibly*! Before 1796 technically means any time whatsoever before
1796 including 1795.

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

Discuss please...

Ask yourself what message you would print out if you were the programmer.
Those error messages are there to be taken seriously or to be treated as
hints and ignored.

Bob

[Lesley Robertson]

"singhals" <[email protected]> wrote in message
news:[email protected]...

J. Hugh Sullivan wrote:

I am posting this to soc.genealogy.methods also because it's the
proper place. But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.


I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

Discuss please...

Hugh


I've had the same problem in PAF and FTM. My work-around is, don't run
the problem report. (g)

I'm afaid that this is also my solution. Or not to worry when you know why

the error has been generated - the problem report is only a way of avoiding
surprises. When you have an index showing several people with the same
forename and surname, it helps a great deal to have an estimate, if only to
let you pick the right one from the index.

Mind you, I think, if I trusted the other researcher who'd given me a
wedding date of 1724, I'd move the birth back to BEF [1724-16] or so.......
Lesley Robertson

[Hugh Watkins]

"J. Hugh Sullivan" <[email protected]> wrote in message
news:[email protected]...

I am posting this to soc.genealogy.methods also because it's the
proper place. But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.


I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

your researcher probably has another person of the same name
if "abt 1724" it is guesswork
because if no ddmmyyyy date is available there has not been reference to
primary sources

It is best stick to one program and learn it thoroughly
I swear by Family Tree Maker and never import other peoples gedcom
so any mistakes and assumptions are all mine and consistent

If in doubt leave it out

Another instance, A person died in 1796 so I list marriage as "Bef
1796 << is a waste of your time and it is better to leave the field empty

they may never have married or have had many wives or changed their name or
the records are lost in a burnt courthouse - many facts are lost forever.

I no longer write "unknown" in a name field for a missing name, but use a
"-" to make the empty field display in a tree

If in doubt leave it out - or put it in the notes

Hugh W

[J. Hugh Sullivan]

On Tue, 12 Oct 2004 16:00:59 GMT, Robert Heiling <[email protected]>
wrote:

"J. Hugh Sullivan" wrote:

I am posting this to soc.genealogy.methods also because it's the
proper place.

If you double-check the respective charters, you'll find that computing is
the right place. I note that methods gets a lot of posts that don't belong
there.

Understood - I find more that I can use on this group anyhow.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755".

Just as an aside, "Bef 1755" may not be correct as he could have been born
in 1755 and still have been 45+. I'd use Abt 1755.

Agreed.

From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Wasn't he?<g

Doesn't it depend on whether he was the father or son?

They could more accurately haver said "possibly born after
marriage" as Bef 1755 could mean sometime before the Abt 1724.

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

Well *possibly*! Before 1796 technically means any time whatsoever before
1796 including 1795.

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

Discuss please...

Ask yourself what message you would print out if you were the programmer.
Those error messages are there to be taken seriously or to be treated as
hints and ignored.

Bob

My problem is that, without checking I can't determine which should be
checked and which ignored. If one assumes all the like-worded problems
are equally superficial, I have very few problems.

One report is much better reporting problems than the other.
Unfortunately it's 46 pages compared to 5. But the 46 page report
takes 5 lines to print what the other does on 1 line and I see no way
to change that unless I edit every line.

Thanks for the comments - you were one of the ones I was looking for.

Hugh

[Robert Heiling]

"J. Hugh Sullivan" wrote:

On Tue, 12 Oct 2004 16:00:59 GMT, Robert Heiling <[email protected]
wrote:

"J. Hugh Sullivan" wrote:
snip
For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755".
snip
From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Wasn't he?<g

Doesn't it depend on whether he was the father or son?

No. It's an error on thee face of it. A person marries in 1724 and then in
1755 decides to be retroactively born?<g>

Bob

[J. Hugh Sullivan]

On Tue, 12 Oct 2004 12:00:36 -0400, singhals <[email protected]>
wrote:

J. Hugh Sullivan wrote:

I am posting this to soc.genealogy.methods also because it's the
proper place. But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.


I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

Discuss please...

Hugh


I've had the same problem in PAF and FTM. My work-around is, don't run
the problem report. (g)

NOW you tell me!

Seriously, if date fields would take a less-than or greater-than or the
approximately-equal symbols, it would help. At least, the formula would
function with them. I don't myself see why the programmer
couldn't/didn't/won't "interpret" bef to less-than and aft to
greater-than and be done with it, but it doesn't seem to have occurred
to them.

Cheryl

My first thought is that it doesn't make any difference whether a
person is "83 years old when married" - maybe I need to look at
unchecking a search for that particular item if such exists.

Representatives of both programs read this news group - if they give
much credence to the problem I expect they will address it, be it
programming or entering data especially when taking 5 lines to print
what could be printed on 2 or less.

Problems like this require the thought of both parties - the
programmers can't program for everything and we need to strive for
consistency in what we record to make their job workable.

When I have a problem I like to have a suggested solution but I
haven't given this sufficient study. My first thought is to start over
and have a plan in mind which overcomes the deficiencies in the
problem reports. But sometimes it's too late to start over.

Another question, Cheryl: does it really matter when a person is
married if we have no idea about the date? We can't presume that they
were married before the birth of children - not even that they were
actually married before death except for common law.

While I have the soap box, one program provides for listing a source
for "name". The other program does a much better job of listing
"events without sources" but does not include the name (because it
doesn't provide for a "name" source. It's sorta minor but if I am the
"world's best clerk and picker of nits" I don't want to miss anything
or be inconsistent.

Your thoughts - or anybody's?

Hugh

Hugh

[Dave Hinz]

On Tue, 12 Oct 2004 16:29:37 GMT, J. Hugh Sullivan <[email protected]> wrote:

On Tue, 12 Oct 2004 16:00:59 GMT, Robert Heiling <[email protected]
wrote:

Just as an aside, "Bef 1755" may not be correct as he could have been born
in 1755 and still have been 45+. I'd use Abt 1755.

Agreed.

Right. But for death dates, I have a probate record of December 4, 1397,
so there it's "bef 4 Dec 1397" for instance, right? I mean, I know he
died a bit before then,and it's very likely that he died in 1397, but
I don't _know_ for sure. By giving the date of the probate record,
and putting the link and a copy of the record in the notes, it'll be
as clear as it can.

Or, would it be better to put "about 1397"? I'm thinking not, but not
sure. Either way, details go in the notes.

From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Wasn't he?<g

Doesn't it depend on whether he was the father or son?

They could more accurately haver said "possibly born after
marriage" as Bef 1755 could mean sometime before the Abt 1724.

Ah, so the date comparison formula looks at year, and doesn't grog "bef"
as anything useful. For this guy, if you have a marriage date of
"abt 1724", wouldn't it make sense to modify his birth date to be "bef 1724"
rather than "bef 1755"? Assuming, as you say, it's the same guy. And,
assuming the "abt" is something that's better than just a SWAG.

"bef" and "aft" imply a hard piece of data is known, "abt" to me says
"guess". Not sure it's appropriate to use a guess to get anything more than
another guess, so maybe this guy's birth would be more of a "abt 1700"?
It's all fuzzy. But put your thinking in the notes either way, so you
can figure out what the hell you were thinking years from now when you
find something else about the guy - that's what I'm doing anyway.

My problem is that, without checking I can't determine which should be
checked and which ignored. If one assumes all the like-worded problems
are equally superficial, I have very few problems.

Does changing a "bef" to an "abt" and adjusting the numeric value help?
Or, does whatever tool you're using allow you to say "Yes, I know, never show
me this particular error for this person again"? Kind of a "I've acknowledged
this concern, just show me the new ones" kind of thing?

If you've got a unix-ish box, you could take a report, keep those lines
from it that you don't want to see, and use 'diff' (or grep -v) to remove
those lines from the resultant report so you'd just see the ones not on your
ignore list...more than a few ways to do it.

Dave Hinz

[J. Hugh Sullivan]

On 12 Oct 2004 17:30:02 GMT, Dave Hinz <[email protected]> wrote:

On Tue, 12 Oct 2004 16:29:37 GMT, J. Hugh Sullivan <[email protected]> wrote:
On Tue, 12 Oct 2004 16:00:59 GMT, Robert Heiling <[email protected]
wrote:

Just as an aside, "Bef 1755" may not be correct as he could have been born
in 1755 and still have been 45+. I'd use Abt 1755.

Agreed.

Right. But for death dates, I have a probate record of December 4, 1397,
so there it's "bef 4 Dec 1397" for instance, right? I mean, I know he
died a bit before then,and it's very likely that he died in 1397, but
I don't _know_ for sure. By giving the date of the probate record,
and putting the link and a copy of the record in the notes, it'll be
as clear as it can.

Or, would it be better to put "about 1397"? I'm thinking not, but not
sure. Either way, details go in the notes.

From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Wasn't he?<g

Doesn't it depend on whether he was the father or son?

They could more accurately haver said "possibly born after
marriage" as Bef 1755 could mean sometime before the Abt 1724.

Ah, so the date comparison formula looks at year, and doesn't grog "bef"
as anything useful. For this guy, if you have a marriage date of
"abt 1724", wouldn't it make sense to modify his birth date to be "bef 1724"
rather than "bef 1755"? Assuming, as you say, it's the same guy. And,
assuming the "abt" is something that's better than just a SWAG.

"bef" and "aft" imply a hard piece of data is known, "abt" to me says
"guess". Not sure it's appropriate to use a guess to get anything more than
another guess, so maybe this guy's birth would be more of a "abt 1700"?
It's all fuzzy. But put your thinking in the notes either way, so you
can figure out what the hell you were thinking years from now when you
find something else about the guy - that's what I'm doing anyway.

My problem is that, without checking I can't determine which should be
checked and which ignored. If one assumes all the like-worded problems
are equally superficial, I have very few problems.

Does changing a "bef" to an "abt" and adjusting the numeric value help?
Or, does whatever tool you're using allow you to say "Yes, I know, never show
me this particular error for this person again"? Kind of a "I've acknowledged
this concern, just show me the new ones" kind of thing?

If you've got a unix-ish box, you could take a report, keep those lines
from it that you don't want to see, and use 'diff' (or grep -v) to remove
those lines from the resultant report so you'd just see the ones not on your
ignore list...more than a few ways to do it.

Dave Hinz



Good hints for a person just starting. And I had not thought of bef,

aft and abt being considered as you say. I wonder how many other times
I have used the terms very casually.

In problem cases like the above perhaps a marriage dat is not that
significant. It's nice to know the fact (if it is) but what difference
does it make in the long run except in a few instances such as
multiple marriages and which children were who's?

At least one program does allow for omitting the prolem in future
reports.

Interestingly enough Genbox does a pretty fair job on the problem
report but it is not quite as thorough.

I don't think I can short cut my effort. I think I've got to look at
each of the 255 problems and make some decision how to use the three
terms logically and consistently. You've helped.

Hugh

[J. Hugh Sullivan]

On Tue, 12 Oct 2004 16:19:33 -0000, "Hugh Watkins"
<[email protected]> wrote:

"J. Hugh Sullivan" <[email protected]> wrote in message
news:[email protected]...
I am posting this to soc.genealogy.methods also because it's the
proper place. But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.


I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.


your researcher probably has another person of the same name
if "abt 1724" it is guesswork
because if no ddmmyyyy date is available there has not been reference to
primary sources

Some of the data is from others and some is my entry. One of my
sources is "Logic/Estimate". I think that is preferable to a note.

It is best stick to one program and learn it thoroughly
I swear by Family Tree Maker and never import other peoples gedcom
so any mistakes and assumptions are all mine and consistent

Over the years several programs have attracted my interest - even now
I switch for specific purposes. However I only export gedcoms from 1
program. And I never import gedcoms except my own. If anything I print
them out and do the manual data entry.

If in doubt leave it out

Another instance, A person died in 1796 so I list marriage as "Bef
1796 << is a waste of your time and it is better to leave the field empty

I am about to conclude the same thing. Entries like that are
essentially from when I first started and were an attempt to be
thorough.

they may never have married or have had many wives or changed their name or
the records are lost in a burnt courthouse - many facts are lost forever.

I no longer write "unknown" in a name field for a missing name, but use a
"-" to make the empty field display in a tree

I insert "??? ???" when there are children but make no entry for the
unknown when there are no children.

If in doubt leave it out - or put it in the notes

Are you saying that you don't use "Bef", "Aft" and "Abt"?

Hugh

[Dave Hinz]

On Tue, 12 Oct 2004 19:10:49 GMT, J. Hugh Sullivan <[email protected]> wrote:

I don't think I can short cut my effort. I think I've got to look at
each of the 255 problems and make some decision how to use the three
terms logically and consistently. You've helped.

Glad to have helped. We hashed this one around quite a bit here maybe
5 years ago or so, I think the gist was to give a hard date when
possible, and make it clear that it's a guess when it is. Probably in one
of the interminable IGI estimated-dates threads.

Dave

[J. Hugh Sullivan]

On 12 Oct 2004 19:27:30 GMT, Dave Hinz <[email protected]> wrote:

On Tue, 12 Oct 2004 19:10:49 GMT, J. Hugh Sullivan <[email protected]> wrote:

I don't think I can short cut my effort. I think I've got to look at
each of the 255 problems and make some decision how to use the three
terms logically and consistently. You've helped.

Glad to have helped. We hashed this one around quite a bit here maybe
5 years ago or so, I think the gist was to give a hard date when
possible, and make it clear that it's a guess when it is. Probably in one
of the interminable IGI estimated-dates threads.

Dave


I've been through the first 60 problems and 58 were "before" dates for

marriage based on the death of an individual. All except one were from
other sources. It appears that more than 90% of the rest are the same
thing and I will delete them also since they are meaningless.

As I view the issue now, I had no logical reason to create or copy
such a date.

Once I correct all this, I'll do another problem report for each
program and see how they look - should be very manageable.

Hugh

[Dave Hinz]

On Tue, 12 Oct 2004 21:00:00 GMT, J. Hugh Sullivan <[email protected]> wrote:

I've been through the first 60 problems and 58 were "before" dates for
marriage based on the death of an individual. All except one were from
other sources. It appears that more than 90% of the rest are the same
thing and I will delete them also since they are meaningless.

Weeellllll, do you know how they were arrived at in the first place?
If they're guesses, OK, but if it's "before 1901" because you have
a probate record or something, then it's still good data. I'd note it,
but that's just me.

As I view the issue now, I had no logical reason to create or copy
such a date.
Once I correct all this, I'll do another problem report for each
program and see how they look - should be very manageable.

Reporting is always fun to try to figure out.

Dave

[Hugh Watkins]

"J. Hugh Sullivan" <[email protected]> wrote

Are you saying that you don't use "Bef", "Aft" and "Abt"?

of course I do but within Family Tree Maker 11 they cause no problems

abt = estimated from a census

I only enter marriages if I have hard data from a church book
or civl registration certificate

anything from IGI gete prefixed IGI in the location field


at the mometn Walter Barber *March 6, 1880 +October 12, 1963
has two spouses

Alice A Lapham born abt 1877

Alma (Dot) Lapham born Aft. 1881

probably the same person but UNPROVEN
until I spend money on marriage certificates

========================
on line here
http://familytreemaker.genealogy.com/us ... index.html

http://familytreemaker.genealogy.com/us ... -0171.html

Walter Barber (son of William R Barber and Selina Barber) was born March 6,
1880 in AMW BB of Bristol and Torquay, and died October 12, 1963 in Beach
Court, Sea Point, Cape Town. Crem. 15 Capetown. He married (1) Alma (Dot)
Lapham, daughter of Alfred Lapham and Fanny Bartley Ball. He married (2)
Alice A Lapham, daughter of Alfred Lapham and Fanny Bartley Ball.

Children of Walter Barber and Alma (Dot) Lapham are:
Molly Barber.

Children of Walter Barber and Alice A Lapham are:
Molly Barber.
Kenneth Barber.

===================================

just found a marriage on ancestry Thomas Buckingham Ball
May 13, 1827 whihc cause an error window when I entered it
because I had a dob as bef 1819

so I changed it to Bef. 1809 which is a pure guess
and now, after this conversation, I have left the dob blank

Hugh W


--
My Blogs

GENEALOGE http://hughw36.blogspot.com/

Exploring Rootsweb http://rootsweb.blogspot.com/
Exploring Ancestry dot com and co uk http://ancestry.blogspot.com/

Getting to know AOL http://aoldotcom.blogspot.com/

Viking http://jorvik.blogspot.com/
1805 http://1805ad.blogspot.com/

Memories http://biog.blogspot.com/
Hugh's Review of Reviews http://hror.blogspot.com/
Hugh's Bloog http://moc-topsgolb.blogspot.com/

[singhals]

J. Hugh Sullivan wrote:

On Tue, 12 Oct 2004 12:00:36 -0400, singhals <[email protected]
wrote:


J. Hugh Sullivan wrote:


I am posting this to soc.genealogy.methods also because it's the
proper place. But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.


I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

Discuss please...

Hugh


I've had the same problem in PAF and FTM. My work-around is, don't run
the problem report. (g)


NOW you tell me!

Hey, no-problemo, amigo. (g)

Seriously, if date fields would take a less-than or greater-than or the
approximately-equal symbols, it would help. At least, the formula would
function with them. I don't myself see why the programmer
couldn't/didn't/won't "interpret" bef to less-than and aft to
greater-than and be done with it, but it doesn't seem to have occurred
to them.

Cheryl


My first thought is that it doesn't make any difference whether a
person is "83 years old when married" - maybe I need to look at
unchecking a search for that particular item if such exists.

No, I don't see a problem in how far over 21 either party is. I DO see
a problem, sometimes, with having a couple marry when one of 'em is 5
or 6, and yeah, I like it when the program mentions it. (g) I know of a
situation where a girl DID marry at 12 or so, and since her first child
was born when she was 22, the marriage date needed explaining; still,
that sort of dynastic legal maneuvering isn't common enough to need a
special field for it.

Representatives of both programs read this news group - if they give
much credence to the problem I expect they will address it, be it
programming or entering data especially when taking 5 lines to print
what could be printed on 2 or less.

Maybe my pain pill kicked in too soon, but -- what? (g) how would
USING "abt" in the calculations enlarge the output? If I subtract 21
from "about 1789" I get "about 1768" (I think -- the important part is
abt gets added to the answer because it was in the equation).

Problems like this require the thought of both parties - the
programmers can't program for everything and we need to strive for
consistency in what we record to make their job workable.

If ONLY the records and our ancestors had done their bit toward consistency.

When I have a problem I like to have a suggested solution but I
haven't given this sufficient study. My first thought is to start over
and have a plan in mind which overcomes the deficiencies in the
problem reports. But sometimes it's too late to start over.

Another question, Cheryl: does it really matter when a person is
married if we have no idea about the date? We can't presume that they
were married before the birth of children - not even that they were
actually married before death except for common law.

We'll never even try to find the date if we don't make some assumptions.
Since you count better'n I do, it may not bother you, but the routine
of checking and re-estimating a possible marriage date everytime I look
at the data gets old for me. So, I use probabilities -- If I find a
death record for Madelaine GotRocks, wife of David GotRocks, I get to
assume there was a marriage, and that it occurred BEF the date of the
death record (I also get to assume that David outlived her); if the
death was reported by Tommy Gotrocks, son, I get to assume Tommy belongs
to Maddy and Dave...and since Tommy has to be 18 or so to get the
responsibility of reporting it, then the marriage occurred BEF
(deathdate minus 18), which narrows my field-of-search some. I mean,
bad enough to have to search 4 states for it, no point in searching 40
years in each state!

While I have the soap box, one program provides for listing a source
for "name". The other program does a much better job of listing
"events without sources" but does not include the name (because it
doesn't provide for a "name" source. It's sorta minor but if I am the
"world's best clerk and picker of nits" I don't want to miss anything
or be inconsistent.

Ummm, would an EVENT necessarily have a name attached? One of my local
throw-aways prints police reports of the " Thursday unknown person or
persons entered a dwelling at 414 Fourty-Fourth between 7pm and 9:15 pm"
Decidedly an event, but no name in sight. ?

Cheryl

[J. Hugh Sullivan]

On 12 Oct 2004 21:18:02 GMT, Dave Hinz <[email protected]> wrote:

On Tue, 12 Oct 2004 21:00:00 GMT, J. Hugh Sullivan <[email protected]> wrote:

I've been through the first 60 problems and 58 were "before" dates for
marriage based on the death of an individual. All except one were from
other sources. It appears that more than 90% of the rest are the same
thing and I will delete them also since they are meaningless.

Weeellllll, do you know how they were arrived at in the first place?
If they're guesses, OK, but if it's "before 1901" because you have
a probate record or something, then it's still good data. I'd note it,
but that's just me.

It appears to me that several people didn't have a marriage date for
some couples. If they were both dead - one in 1991 and the other in
1998 for example - a marriage date was entered as "bef 1991". That's
obvious (if they were actually married and not in 1991) but
superfluous. It looks like one of my programs picked up more than 200
of those. I've got 195 more of them to review!

Hugh

[Dave Hinz]

On Wed, 13 Oct 2004 13:19:21 GMT, J. Hugh Sullivan <[email protected]> wrote:

On 12 Oct 2004 21:18:02 GMT, Dave Hinz <[email protected]> wrote:

Weeellllll, do you know how they were arrived at in the first place?
If they're guesses, OK, but if it's "before 1901" because you have
a probate record or something, then it's still good data. I'd note it,
but that's just me.

It appears to me that several people didn't have a marriage date for
some couples. If they were both dead - one in 1991 and the other in
1998 for example - a marriage date was entered as "bef 1991". That's
obvious (if they were actually married and not in 1991) but
superfluous. It looks like one of my programs picked up more than 200
of those. I've got 195 more of them to review!

Ah, right. Sounds like you're being careful about it; never hurts to clean
up when it's done well. Yes, that's an obvious "no kidding?" kind of a
non-data-piece, that adds no value and confuses problem reports.
If they had kids, you could always pick an "abt" for the year before
the first kid was born; that'd get it in the right decade maybe and is
certainly more useful than "before they died". It's an assumption, but
that's why it's an "abt".

Dave Hinz

[J. Hugh Sullivan]

On Wed, 13 Oct 2004 09:03:47 -0400, singhals <[email protected]>
wrote:

We'll never even try to find the date if we don't make some assumptions.
Since you count better'n I do, it may not bother you, but the routine
of checking and re-estimating a possible marriage date everytime I look
at the data gets old for me. So, I use probabilities -- If I find a
death record for Madelaine GotRocks, wife of David GotRocks, I get to
assume there was a marriage, and that it occurred BEF the date of the
death record (I also get to assume that David outlived her); if the
death was reported by Tommy Gotrocks, son, I get to assume Tommy belongs
to Maddy and Dave...and since Tommy has to be 18 or so to get the
responsibility of reporting it, then the marriage occurred BEF
(deathdate minus 18), which narrows my field-of-search some. I mean,
bad enough to have to search 4 states for it, no point in searching 40
years in each state!

I do a lot of "what ifs" looking for ancestors before 1816. I assume
marriage at 21 for the male and 18 for the female; I assume 33 years
per generation and 2 years between births for each child. Although not
very precise I'm surprised at several things: (1) it eliminates a lot
of people who fall too far outside the range (like your no 40 year
search) (2) It points out a number of families who fit the pattern
extremely well once you know who to look for and learn the actual
facts.

However knowing that a couple were in their 70s when they passed , one
passed in 1840 and the other in 1845 and using "bef 1840 for a
marriage date accomplishes very little.

I used to also assume that a property owner had to be 21 - that was
not absolute. Primogeniture was also not absolute.

There is no escape from caution.

While I have the soap box, one program provides for listing a source
for "name". The other program does a much better job of listing
"events without sources" but does not include the name (because it
doesn't provide for a "name" source. It's sorta minor but if I am the
"world's best clerk and picker of nits" I don't want to miss anything
or be inconsistent.

Ummm, would an EVENT necessarily have a name attached? One of my local
throw-aways prints police reports of the " Thursday unknown person or
persons entered a dwelling at 414 Fourty-Fourth between 7pm and 9:15 pm"
Decidedly an event, but no name in sight. ?

Cheryl

I mistated - I wanted the program to list names of people in my data

base where I did not list my source for the name - cause I got it from
sumwhere. The program calls that an "event".

Hugh

[Robert Heiling]

Robert Heiling wrote:

"J. Hugh Sullivan" wrote:

On Tue, 12 Oct 2004 16:00:59 GMT, Robert Heiling <[email protected]
wrote:

"J. Hugh Sullivan" wrote:
snip
For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755".
snip
From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Wasn't he?<g

Doesn't it depend on whether he was the father or son?

No. It's an error on the face of it. A person marries in 1724 and then in
1755 decides to be retroactively born?<g

Hugh! You didn't comment on this and I'm still sort of sitting here
scratching my head regarding the above.

My assumption when you first presented this problem was that the Birth & the
Marriage were for the same person. Other posts on the thread seem to confirm
that. That's why your "father or son" mention confuses me. If the Marriage is
for the father & the Birth for the son, then it's a different ballgame.

But taking it as the single person scenario, Abt 1724 is much like just saying
1724. Abt in it's +/- is fairly subjective and I haven't seen any standard
definition. TMG has a default of +/-3 years, but gives the option to change
that, but how would anyone else know what you meant?<g> Bef 1755 means just
that and goes backwards from 1754 to the beginning of time.

What that phase of the audit program would do is pick various dates and run
them through some comparison tests. It has to take Abt 1724 at the face value
of 1724 Marriage and compare with 1754 Birth which is where your error message
gets generated. It would continue to go back to 1724 and continue on. Other
values it runs past such as 1720 Birth 1724 Marriage don't warrant another
error "marriage at too young age" or somesuch since the serious of the first
message is considered greater. Either that or, having already issued one error
msg, it leaves it go at that. It's probably the latter.

Bob

[J. Hugh Sullivan]

On Wed, 13 Oct 2004 21:29:15 GMT, Robert Heiling <[email protected]>
wrote:

"J. Hugh Sullivan" wrote:

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755".

From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Wasn't he?<g

Doesn't it depend on whether he was the father or son?

No. It's an error on the face of it. A person marries in 1724 and then in
1755 decides to be retroactively born?<g

Hugh! You didn't comment on this and I'm still sort of sitting here
scratching my head regarding the above.

Sorry about that.

I picked up the father's name from some researcher. I later found him
in the 1800 census and I made the entry "bef 1755" for birth date
because of the "45+" entry on the 1800 census. At the time I saw no
harm. Later I got the "abt 1724" for his marriage date but didn't
change the "bef 1755" date. Probably I should have changed it to "abt
1700".

I gather that the Problem Report pays no attention to the "bef" part
of the entry so 1755 being later than 1724 a problem was reported.

I made a decision to delete all entries of the type "bef 1755" and
decided not to input "abt 1700". I think there were about 225 of them
and I have deleted about half.

I think it's better to leave the abt. birth date blank unless and
until I research the family. At that time I would probablt fill in
some tentative dates for research purposes.

Seems to me like I ought to establish some rule of thumb for
approximations. Before and After should probably be used sparingly;
About should be used when there is some basis of fact for reference.

In the above example I'm not sure "abt 1724" is a sufficient reference
to use "abt. 1700" as a birth date although it would probably do no
harm.

Unfortunately I'm noticing that I picked up a number of dates from
other researchers with "Between" with a 20 year or better range for
birth and death. That results in a number of reported problems. When I
run my next problem report I'll see what I can do about that. I think
those wide ranges come from FTM.

I've done enough testing to indicate that my marrying ages of 21 and
18, 33years for a generation and children born every two years is at
least as good as the range method. When one is researching it's easy
to expand my dates to a range.

Plus, my data is not on a web site or published and I exchange only on
a direct line basis - so people should not be misled by entries I
make. Unfortunately an old theory or two, marked clearly as such, has
been adopted by a few people as gospel. Fortunately they didn't give
me credit because I have since discarded those theories.

Between all the help here and my own efforts I have learned a lot.

Hugh
Hugh

[Robert Heiling]

"J. Hugh Sullivan" wrote:

On Wed, 13 Oct 2004 21:29:15 GMT, Robert Heiling <[email protected]
wrote:

"J. Hugh Sullivan" wrote:

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755".

From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Wasn't he?<g

Doesn't it depend on whether he was the father or son?

No. It's an error on the face of it. A person marries in 1724 and then in
1755 decides to be retroactively born?<g

Hugh! You didn't comment on this and I'm still sort of sitting here
scratching my head regarding the above.

Sorry about that.

I picked up the father's name from some researcher. I later found him
in the 1800 census and I made the entry "bef 1755" for birth date
because of the "45+" entry on the 1800 census. At the time I saw no
harm. Later I got the "abt 1724" for his marriage date but didn't
change the "bef 1755" date. Probably I should have changed it to "abt
1700".

Ummm. From a genealogy point of view you should have started wondering about
how many people you were (and still are?) dealing with.

I gather that the Problem Report pays no attention to the "bef" part
of the entry so 1755 being later than 1724 a problem was reported.

There is a zone there that would be a problem. I don't know as a hard fact just
how that audit works since I didn't program it, but it's certainly a reasonable
conjecture on their algorithm.

I made a decision to delete all entries of the type "bef 1755" and
decided not to input "abt 1700". I think there were about 225 of them
and I have deleted about half.

I think it's better to leave the abt. birth date blank unless and
until I research the family. At that time I would probablt fill in
some tentative dates for research purposes.

I think Abt is a useful tool and I use it accompanied by a note as to my reason.

Seems to me like I ought to establish some rule of thumb for
approximations. Before and After should probably be used sparingly;
About should be used when there is some basis of fact for reference.

In the above example I'm not sure "abt 1724" is a sufficient reference
to use "abt. 1700" as a birth date although it would probably do no
harm.

It was a pretty outlandish example and it shouldn't stop you from using those.
I'd like to see some of the others before agreeing with you on the decision to
stop.

Unfortunately I'm noticing that I picked up a number of dates from
other researchers with "Between" with a 20 year or better range for
birth and death. That results in a number of reported problems. When I
run my next problem report I'll see what I can do about that. I think
those wide ranges come from FTM.

Beware of those researchers!<g> What was the Source for that 1724 for a guy you
already knew was born ~1755???? Something wrong with that picture.

I've done enough testing to indicate that my marrying ages of 21 and
18, 33years for a generation and children born every two years is at
least as good as the range method. When one is researching it's easy
to expand my dates to a range.

Sounds good to me.

Plus, my data is not on a web site or published and I exchange only on
a direct line basis - so people should not be misled by entries I
make. Unfortunately an old theory or two, marked clearly as such, has
been adopted by a few people as gospel. Fortunately they didn't give
me credit because I have since discarded those theories.

Between all the help here and my own efforts I have learned a lot.

It's all interesting, isn't it?

Bob

[J. Hugh Sullivan]

On Thu, 14 Oct 2004 00:30:48 GMT, Robert Heiling <[email protected]>
wrote:

"J. Hugh Sullivan" wrote:

In the above example I'm not sure "abt 1724" is a sufficient reference
to use "abt. 1700" as a birth date although it would probably do no
harm.

It was a pretty outlandish example and it shouldn't stop you from using those.
I'd like to see some of the others before agreeing with you on the decision to
stop.

Maybe it would be better said that I plan to make more judicious use
of the terms after reviewing each particular situation.

My example was a bit outlandish but it was typical of the problems I
had because I had not looked carefully enough at data furnished by
others - I merely copied if there was no apparent (to me) conflict.

During our first few years in genealogy I think we periodically change
our methods and parameters which results in inconsistencies. By the
time we have a few thousand names we forget to review what we have
done compared to what we do now.

Hugh

[D. Stussy]

On Tue, 12 Oct 2004, J. Hugh Sullivan wrote:

I am posting this to soc.genealogy.methods also because it's the
proper place. But I think some people I want to hear from are not on
that news group. Please excuse if you see the post twice. Depending on
the response I will be more selective when I respond.


I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Why do you think that this ISN'T a problem?

If he married about 1724, it should be obvious from THAT piece of data that he
should have been born before 1708 (assuming age 16 at marriage), so
technically, the "before 1755" birthdate is inconsistent, in that it allows for
a birthdate of the range 1724-1754 which doesn't fit the marriage fact.
Although that birth is a true "fact," there is a better interpretation of the
fact: Before 1708, consistent with BOTH.

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

Age at marriage is a problem only if they are too YOUNG. There should be no
upper limit set. Why is that even a problem?

I'm sure there are other equally troublesome "problems".

Seems like "Bef", "Abt" and "Aft" create as many problems as they
solve especially if use with marriage.

In both cases above I'm thinking that my decision to use "Bef' dates
was unnecessary and even misleading.

Discuss please...

Hugh

[J. Hugh Sullivan]

On Mon, 18 Oct 2004 00:34:31 GMT, "D. Stussy"
<[email protected]> wrote:

On Tue, 12 Oct 2004, J. Hugh Sullivan wrote:

I printed a list of potential problems in one genealogy data base
using both Legacy and RootsMagic.

I see some peculiarities which I need to solve.

For example a census shows a person as "45+" in 1800 so I listed the
birth date as "Bef 1755". From a researcher I have a marriage date as
"Abt 1724". The problem report shows "born after marriage".

Why do you think that this ISN'T a problem?

If he married about 1724, it should be obvious from THAT piece of data that he
should have been born before 1708 (assuming age 16 at marriage), so
technically, the "before 1755" birthdate is inconsistent, in that it allows for
a birthdate of the range 1724-1754 which doesn't fit the marriage fact.
Although that birth is a true "fact," there is a better interpretation of the
fact: Before 1708, consistent with BOTH.

There is a better interpretation if the 1724 date is correct. I have
to decide whether to use the 1724 date. If I do I'll use 'Abt 1703" as
the DoB since I use 21 as the average for men to marry. But I will
also look at the age of the female and the source of my source (the
person who gave me the info) for the date.

Another instance, A person died in 1796 so I list marriage as "Bef
1796" and the report shows "73 years old when married".

Age at marriage is a problem only if they are too YOUNG. There should be no
upper limit set. Why is that even a problem?

The problem report in 2 genealogy programs reports it as a potential
problem. I solved it by telling the programs to ignore the problem, by
individual, after I studied each case. In a couple of cases the
problem was real vice potential.

I am concerned about two areas: (1) the reliability of date furnished
me by others (2) my parameters when I entered data. As we gain
knowledge in genealogy our methods change.

Hugh