* Research analysis and conclusions

[natashahouseman]

Hi folks,

I am seeking some insight into how people use family historian notes/research notes or other approaches to record analysis and conclusions.
I am using Family Historian as a route to improving the rigour of my family history research by starting my tree from scratch rather than importing a gedcom. With 500+ direct ancestors identified this is a long term project and so I am trying to work systematically and think carefully up front about how I use family historian to it's best ability. I started with sources and citations (having never sourced or cited beyond linking on ancestry or thinking "baptism" was sufficent) and with support from this group I think I have worked out my approach to those.
I am now moving onto how I record analysis and conclusions. I have been lucky as in most cases as although it might have taken a while to figure out the connections once I've then identified and added the sources, the conclusions are mostly self-explanatory, but occaisionally they are not
I am thinking at the moment a short note on the individual to explain the situation eg "There are two Henry Claphams born in the parish in the relevant period but there is evidence to suggest the other one died young" and then an adapted research note explaining in more detail the analysis. Then the quick note pops up in the narrative and the longer research note is there to go back to. However I don't think it's possible to link sources which i have consulted which have proved the negative. I was also thinking about adding the sources to eg the birth with a citation note as to why they prove the negative. But it's all feeling a bit messy and not systematic.
I wonder if anyone might be willing to explain how they use notes, research notes or other approaches to record conclusions to help me think through how I approach this.
Many thanks
Natasha

[AdrianBruce]

For an extended explanation such as you describe, I simply use a Note Record and link it to as many things (individual records, family records, source records, etc?) as seems useful.

I don't use a Research Note for that purpose for the simple reason that I was creating these long before Research Notes came in. I also try to format the top of these conclusion note records in the same fashion so that they sort together. So at the moment, the first line of my conclusion note record looks like (say)

[[ Genealogical notes - Frank Taylor - Identification of his wife

and the last line is (not surprisingly)
]]

The [[ ]] are there to inhibit the same extensive text appearing against (say) Frank, his wife and each of their children, each of whom might be mentioned, so the same Note Record might be linked to each (I use narrative Reports - not because I think their text is particularly readable but because it's a quick way of getting something). The same phrase "Genealogical notes - " appears at the start of all such note records. Probably a quick summary such as you suggest, outside [[ ]] brackets, might be useful in some circumstances - not for Frank's wife, because that's just a conclusion, but certainly for Frank himself, who was illegitimate and whose birthplace varied in the censuses. Though actually in his specific case, the specific note appears as part of the note against his birth event.

As I implied above, the [[ Genealogical notes ... ]] can also be linked to a number of Source Records to justify why this source record for Frank Taylor and this one all refer to the same Frank and contribute to the end conclusion. I'm very careful about proof of identification in sources - having a source record for Frank Taylor isn't the point - it's how do I know it's my Frank Taylor that's important. Occasionally I need to link several source records together before I can confirm who it is - linking the Note Record to each Source Record is an important part of that.

Negative Records - I usually just describe each negative record in plain text - for instance, "No Theophilus P Vildebeest appears in FreeBMD".

Citations of Sources in the [[ Genealogical notes ... ]] record - I guess I could, but I tend to just describe the thing in English - after all, the formal citation will be against the fact, so why have it all in two places? (Otherwise, it all gets a bit chicken and egg)

[natashahouseman]

Many thanks Adrian. I hadn't come across [[ ]] as a tool (lots of learning to do) and I will experiment with it now as I think that might give me what I need, as I like you, think the narrartive reports are clunky, but do a good job in bringing together all the information I have about a person.
Still pondering on the negative information. I am definitely with you on the "searched freebmd with no results" but sometimes I end up ordering 2 or 3 certificates to help solve a problem and I'd like to include them as a source, but then maybe they are a source that is only connected to the note and not to the individuals who are involved.

[ColeValleyGirl]

There is a difference between negative evidence and negative findings that might be worth bearing in mind: Negative Evidence vs. Negative Findings.

And there's a wish list item to include Negative Evidence as an option for the Citation Assessment: Add Values to Citation Assessment (V7).

More generally, in V6 and before I created Word/PDF documents to document how I reached 'complex' conclusions --- see the pdf associated with the baptism here. (It's attached as Media to the Fact).

In V7 I may move to using formatted Notes (not Research Notes, because I want to include the documentation on my website via reports.)

[AdrianBruce]

Irrelevant comment - I am entranced with the occupation of "Sword Blade Forger"

[ColeValleyGirl]

Irrelevant comment - I am entranced with the occupation of "Sword Blade Forger"

I so wanted him to be my ancestor.

[mjashby]

Obviously there must be a need to watch more niche US TV shows such as "Forged in Fire".

Descriptive Quote: "In each episode, four bladesmiths compete in a three-round elimination contest to forge bladed weapons, with the overall winner receiving $10,000 and the show's championship title "Forged in Fire Champion"."

There was probably good money to be made through the skilled forging of sword blades. And those customers who purchased poor forgeries probably didn't survive long enough to complain.

Mervyn

[KFN]

Having worked with GEDCOM since the 1980s and now participating in discussions about future updates, one of the biggest issues with using GEDCOM as the basis for data storage is the adherence to a more “conclusion” based design rather than a “research” based (data collection) design.

Most software (and GEDCOM) assumes for example one birth fact, which when doing research may not be valuable! The same can be said about marriage data and many other “facts and events”

Relationships between people are hard to make “tentative”!

And heaven forbid you are doing a name study or a rural town or county person collection where you will be looking for relationships over time rather than immediately.

This all being said, I use other databases on my computer, maintain a document catalog/repository to record information until I can make some level of conclusion about a person before entering the information in FH!

[tatewise]

Most software (and GEDCOM) assumes for example one birth fact, which when doing research may not be valuable! The same can be said about marriage data and many other “facts and events”.

Most software and GEDCOM allow multiple instances of any fact including Birth, Marriage, and Death events.

[kfunk_ia]

Most software (and GEDCOM) assumes for example one birth fact, which when doing research may not be valuable! The same can be said about marriage data and many other “facts and events”.

Most software and GEDCOM allow multiple instances of any fact including Birth, Marriage, and Death events.

However, even though most software allows multiples of facts, this doesn't make them overly useful. John Smith may have 4 Birth facts based on various sources, but if one were to look at the fact listing for a person, I would see 4 facts but not necessarily which one is the "best" fact.

My workaround in RM, which will likely follow me to FH (unless I am clearly missing something) is to create alt facts, so I have a 'Birth (Alt)', 'Death (Alt)', and 'Marriage (Alt)' that I use frequently. Birth will contain the most likely or strongest fact and the various alt facts will contain lesser supported dates along with the source of and reasoning for why that is a weaker possibility.

I started doing this before we had research notes or logs built in and I haven't yet explored using those tools to alter my workflow. I still haven't fully wrapped my mind around FH's method of Research Notes at this time.

[tatewise]

In FH you can set Fact Flags to indicate the best or dubious ones, etc.
e.g. Preferred, Tentative, Rejected

So the "best" Birth event would have the Preferred Fact Flag.

Other users prefer to have just one Birth event with the "best" characteristics and include an analysis of all the Source Citations to justify why those characteristics were chosen and others rejected.

[kfunk_ia]

In FH you can set Fact Flags to indicate the best or dubious ones, etc.
e.g. Preferred, Tentative, Rejected

Aye! And that has its own set of issues. For example my 'preferred' birth is now listed as Birth[2], the rejected birth is Birth. That is going to be some crazy fun when writing queries. Then there is the issue with this printing as 'Birth (2)' on my family group sheet.

[jelv]

It's a doodle to write queries!

Code: Select all

%INDI.BIRT[preferred]%

Edit to add: If there is no preferred, or only one it defaults to [1]

[KFN]

In FH you can set Fact Flags to indicate the best or dubious ones, etc.
e.g. Preferred, Tentative, Rejected

So the "best" Birth event would have the Preferred Fact Flag.

Other users prefer to have just one Birth event with the "best" characteristics and include an analysis of all the Source Citations to justify why those characteristics were chosen and others rejected.

But these flags are “conclusions” not “research”! In science you don’t assert a conclusion while doing your research, this is a difference that most people don’t understand! Besides, “Preferred, Tentative, Rejected” are not a GEDCOM characteristic, but 2 are still “conclusions”, all entries would be “tentative” during the research phase.

[jelv]

I have a case where two birth dates are both proven facts with good sources. My grandmother's birth certificate shows her birth date as 22nd May; that is the date she always celebrated her birthday on. However I found her baptism which was on 21st May (the day before she was born!). The baptism helpfully has her birthday as 22nd April. The birth was registered on 4th July. I suspect they were trying to avoid the fine for late registration - but they miscalculated as it is still 43 days after the date they gave the registrar. My only quandary is which of the two birth events I should set as the preferred.

[jelv]

The other thing I haven't checked is whether FH takes notice of the preferred when calculating ages on the Facts tab.

[natashahouseman]

I have a case where two birth dates are both proven facts with good sources. My grandmother's birth certificate shows her birth date as 22nd May; that is the date she always celebrated her birthday on. However I found her baptism which was on 21st May (the day before she was born!). The baptism helpfully has her birthday as 22nd April. The birth was registered on 4th July. I suspect they were trying to avoid the fine for late registration - but they miscalculated as it is still 43 days after the date they gave the registrar. My only quandary is which of the two birth events I should set as the preferred.

I have a similar one! My Grandma always told us that her Dad registered her birthdate on the wrong day (the day after she was born). We always celebrated the earlier date and it's how it's entered in eg her Grandmother's date book (although she has the wrong year!). I think the thing to remember is that the souce of a birth certificate can be wrong - it's very unusual for us to be able to identify this, but I would say that the "correct" birth is likely the earlier one, with the one she celebrated and the one on the register being an alternate rather than both be factually correct.

[natashahouseman]

Just again what to thank Adrian. As I have just resolved another case involving either one or two William Watsons. Most others have one William Watson marrying twice, I think it's two, which I also think leads to the conclusion that the second marriage (to a Martha Watson) was to his cousin.
I've included as shared note and linked it to the different parties and also used this to include on a couple of shared sites including family search. Critically I've been able to link a source to the shared note and the note to the source (circular I realise, but helps ensure I don't have a floating source). In other words, I think this is a very effective solution for me. Just need to decide on a systematic way to name these and what I want to show in reports.
Thank you!

[tatewise]

The other thing I haven't checked is whether FH takes notice of the preferred when calculating ages on the Facts tab.

FH does use the Preferred Birth for calculating Ages.

[AdrianBruce]

...
But these flags are “conclusions” not “research”! In science you don’t assert a conclusion while doing your research, this is a difference that most people don’t understand! ...

But that depends on what you think your events and attributes are. As far as I'm concerned, my events and attributes are my conclusions - never final, of course. Others work differently and write their conclusions up elsewhere.

[KFN]

...
But these flags are “conclusions” not “research”! In science you don’t assert a conclusion while doing your research, this is a difference that most people don’t understand! ...

But that depends on what you think your events and attributes are. As far as I'm concerned, my events and attributes are my conclusions - never final, of course. Others work differently and write their conclusions up elsewhere.

You are missing the point of my reply to the OP. In GEDCOM the design was to provide “conclusions” to LDS for their baptismal to Christ of deceased individuals! They (and most software) don’t do a good just of collecting all data pertaining to an individual and reporting multiple proven (and disproven) facts and events and showing the sources of the non-fact (miss applied information) associated with a person!

The OP wanted to understand how to manage both proven and disproven facts and their sources. GEDCOM does not provide for a way to assert that a source is wrong and creating a citation that this as an incorrect fact.

For example: My wife’s family believes that one child was born 11 months after the supposed biological father died! We know this can’t be true but family stories persist that the biological father is the dead man. GEDCOM and most software does a poor job of reporting this disproven and impossible story, while at the same time asserting the possibility of another as of yet unknown but possible set of fathers (could be one of three)!

[Little.auk]

In FH you can set Fact Flags to indicate the best or dubious ones, etc.
e.g. Preferred, Tentative, Rejected

So the "best" Birth event would have the Preferred Fact Flag.

Other users prefer to have just one Birth event with the "best" characteristics and include an analysis of all the Source Citations to justify why those characteristics were chosen and others rejected.

But these flags are “conclusions” not “research”! In science you don’t assert a conclusion while doing your research, this is a difference that most people don’t understand! Besides, “Preferred, Tentative, Rejected” are not a GEDCOM characteristic, but 2 are still “conclusions”, all entries would be “tentative” during the research phase.

The flags are not "conclusions", but rather "assessments" based on current evidence - "preferred" being the most likely (In science "the current best hypothesis". Even in science there are very few "cast in stone" conclusions).

So, actually, only the "Rejected" flag can be deemed a conclusion, as both "Preferred" and "Tentative" imply varying degrees of uncertainty - which is what we have with all genealogical data. It is not unusual to have several sources for an event all giving different data, and even the original records can be unreliable.

"While doing your research" you are always doing these type of assessments, to direct the route of your ongoing research.

[natashahouseman]

The OP wanted to understand how to manage both proven and disproven facts and their sources. GEDCOM does not provide for a way to assert that a source is wrong and creating a citation that this as an incorrect fact.

For example: My wife’s family believes that one child was born 11 months after the supposed biological father died! We know this can’t be true but family stories persist that the biological father is the dead man. GEDCOM and most software does a poor job of reporting this disproven and impossible story, while at the same time asserting the possibility of another as of yet unknown but possible set of fathers (could be one of three)!

Okay, I think i understand the point now. I'm using FH as a route to checking all my research and properly citing it, so had intended only including individuals (and facts) where I am reasonably certain of the relationship (and fact) (and I have a long long way to go with that having 500+direct ancestors), but at some point i'm going to come across a speculative parent or spouse. It's easy enough to add a note and a range to facts such as birth, but less easy to do this for a relationship to another person. So I need to plan for a method to take account of this. I generally test, gather evidence and work everything out on my ancestry tree first (which is private). But it's those relationships which are more complex to analyse which are the most fun to research so it might be sooner rather than later.

[RS3100]

I add such individuals to a named list, "Unconfirmed Relatives", together with a box expression to show a question mark against those list members in charts and diagrams. A note attached to the individual record explains the evidence that supports or suggests the relationship, and the reasons why it is not conclusive.

By suppressing or editing fact sentences, together with the use of {note} it is also possible to include any doubts and the reasons for them in narrative reports.

[natashahouseman]

I add such individuals to a named list, "Unconfirmed Relatives", together with a box expression to show a question mark against those list members in charts and diagrams. A note attached to the individual record explains the evidence that supports or suggests the relationship, and the reasons why it is not conclusive.

By suppressing or editing fact sentences, together with the use of {note} it is also possible to include any doubts and the reasons for them in narrative reports.

Thanks. I was thinking of something of a list or a flag with report suppression. I am beginning to think i need to write my own user guide for this including formats, decisions, reasoning etc. I can see how amazing FH is going to be for the analysis that ancestry just can't do and I am excited by the potential.