GEN-HISPANIC-L Archives

Archiver > GEN-HISPANIC > 2004-03 > 1079808635


From: (Mike Smith)
Subject: Re: Family Group Sheet Estimation
Date: 20 Mar 2004 10:50:35 -0800
References: <ccd4de67.0403182315.540c1d4e@posting.google.com> <zDBjVzOG5xWAFw88@hyperpeople.demon.co.uk>


> It has nothing to do with genealogy.

==== I refute this. If you are searching for your ancestors
(genealogy)
and you are at the end of a line, where in time do you look? Do you
just give up? :-/

>
> It has everything to do with the statistics of population studies.
>
> It probably does no more than re-invent a wheel already invented and
> long used by official statisticians and actuaries.

This is true, but how do you estimate, without math, events and
happenings? If you have a better suggestion, I am ALL ears. I have no
doubt that an actuary or statistician could do a better job, but the
input I had came from 20 plus people in mathematics, computer
programmer, and family historians. All of these people are +$45K pa
(circa. 1997) in their fields. What makes one "official" is opinion
and not fact and you being a Dr should know this.

>
> It may be worth looking into by such people to see if you have hot on
> anything new which could be used to improve their models.

==== ??? I can't understand the above, but I am sure in a 1Kb program
there is nothing new.

>
> It almost certainly will give rise to a lot of misinformed discussion as
> people seek to do what cannot be done. YOU CANNOT APPLY STATISTICS TO
> SINGLE EVENTS. If you take me as a single event, I wouldn't expect your
> statistical model to be of any value in calculating my life events or

===== I would be interested to know the exact differences and those
circumstances; to suggest that I am not willing to improve the model,
with enough input is stupid. The model fits me to a T EXCEPT you
remove child #1
the dates for child #2-#4 are spot on and it needs to add 2 more kids
to the expected 2.4 kids. I am not looking forward to the year 2030
and praying that medicine improves ;-). We took time to travel Europe
after marriage, to see my results input Male born 1963

> those of my immediate affines. Indeed you only got one thing right for
> me: I married on a 26th October and my wife produced our first child on
> the following 27th August. Furthermore, even where you use statistical

=== Actually I have to giggle here because the human sex drive is such
that from marriage to first born is SO consistant that when the
computer did the results I was expecting a quadratic curve with the
coming of the pill in the 1960s. I expected a straight line with a
hook at the end. The line is so straight, I just simply put in a +1
year.

> words properly, you don't use them correctly about the actual
> populations rightly covered by the newsgroups to which you have posted.

===== Excuse me I was excited about the program, it had 12,000 usages
in 24 hours yesterday after being announced. I am proud of the time
and effort that went into the program and wanted to show it off. While
I was sceptical to show it to everyone I did leave out Africa, the
West Indies, etc I did try, but in the same light I figure that there
is such a diversity of people in the US it is hard to pick the exact
skin color of these people I studied. I know a vast majority were
white, nordic and english if they were related to mormons.

> Even as statistics they are only true of the sample which you studied -
> what is in LDS records. That necessarily does not include a long list of
> other kinds of persons who are resident in and nationals of the
> countries concerned.

==== I would have preferred to study more people but do you know where
there are millions of computerized records in 1997 I could suck, and I
mean suck off a computer, in textual form, so I can hand build a
program around?

>
> Can we now get back to real genealogy?

IMHO this is genealogy and I consider it important.

<snip>
> >I have written a genealogy tool and put it up on the web for all to
> >see and comment on.
> >
> >It can be found at www.ukcs.net/genes
> >
> >I call it Family Estimator.
> >
> >It will take a single date and event for one person and estimate and
> >build a whole family group sheet for them.
> >
> >It comes from over 100,000 computerized genealogical records from
> >Western Europe and North America taken from the LDS databases. The
> >logic is that people marry about 18-24, without birth control, the
> >first child will come 1 year later, and so forth. The logic carries
> >that each person will live an average of X years depending on when
> >they were born. Women will live longer than men. It is for the end of
> >a line. It will give you a point in time to start looking plus or
> >minus 8 yrs. The data is statistical using quadratic and linear
> >equations, so each year and circumstances will produce different data.
> >It takes into account that medicine has improved so has the human life
> >span. It cannot take into account war, plague, illegitimate children,
> >early birth, accidental death or divorce.
> >
> >I would appreciate any comments or queries on this matter.
>

This thread: