GenoPro Home
GenoPro Home  |  Get Started With My Genealogy Tree  |  Buy  |  Login  |  Privacy  |  Search  |  Site Map
 

GenoPro Support Forum




The Uncertainty Principle

Click to view RSS...
Author I think it can be said that no family tree will ever contain the whole truth, but it should as far as possible contain all discovered facts about the ancestors and ideally the sources of those facts.
Posted Tuesday, July 21, 2015 - Post #35255
Legendary Master

Legendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary Master

Administrators
Customers
Important Contributors
FamilyTrees.GenoPro.com
GenoPro version: 3.0.1.0

Last Login: 23 minutes ago
Posts: 2,514, Visits: 17,820
I think it can be said that no family tree will ever contain the whole truth, but it should as far as possible contain all discovered facts about the ancestors and ideally the sources of those facts. 

As appleshaw and others have pointed out, the facts can vary depending on the source of the information, memories fade, errors creep in on transcription, different censuses record the same family member differently, people may deliberated obscure their past. etc.

Therefore we should be able to record all variations in the data discovered.  I suggest the following for GenoProX:

  1. Individuals can have multiple names, not just the 'Alternative Name' available currently but multiple sets for all name fields (forenames, last names, nicknames etc).
    As well as coping with the variations in sources this will also help for example in cases where immigrants from places like Russia and Poland to the UK changed their names to 'Anglicised' versions, also where people have different names in other languages e.g. Hebrew and English.

  2. Similarly dates and places of birth and death can also be recorded differently depending on the source, and so multiple sets of the data associated with birth and death events should be possible.

  3. Sometimes there appears to be more than one possibility in the records for the a parent of a child. I suggest a child can 'belong' to more than one family or alternatively more than one father or mother can be attached to a family unit.

  4. In all the above there should be a way of expressing a 'level of certainty' against each item in the multiple set. In the case of point 3 above this could be indicated visually in the pedigree link by moving from a solid line to a more and more dashed then dotted line as the level of uncertainty rises.



'lego audio video erro ergo disco' or "I read, I listen, I watch, I make mistakes, therefore I learn"


Edited: Tuesday, September 29, 2015 by GenoProSupport
Posted Tuesday, July 21, 2015 - Post #35256
Legendary Master

Legendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary Master

Administrators
Moderators
Customers
Gamma
Translator
GenoPro version: 3.0.0.7

Last Login: 2 days ago @ 12:03 PM
Posts: 4,365, Visits: 21,495
So would you like each object, event, record to have a field name "Certainty"?  By the default, this field would be blank, and people could pick values: How many levels of certainty?  At the moment, GenoPro 2016 uses the following for sources,  however this certainty could be for every object:
  • Unreliable evidence or estimated data
  • Questionable reliability of evidence (interviews, census, oral genealogies, or potential for bias for example, an autobiography)
  • Secondary evidence, data officially recorded sometime after event
  • Direct and primary evidence used, or by dominance of the evidence
Posted Tuesday, July 21, 2015 - Post #35261
Legendary Master

Legendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary Master

Administrators
Customers
Important Contributors
FamilyTrees.GenoPro.com
GenoPro version: 3.0.1.0

Last Login: 23 minutes ago
Posts: 2,514, Visits: 17,820
Every event, relationship and pedigree link should have a way to indicate degree of certainty or confidence level in the accuracy of the information. Where there is a choice of possible values the a way of expressing the relative levels between the choices. 

The examples from GenoPro are taken directly from Gedcom's QUAY (quality) tag:

CERTAINTY_ASSESSMENT: = {Size=1:1} [ 0 | 1 | 2 | 3 ] 
The QUAY tag's value conveys the submitter's quantitative evaluation of the credibility of a piece of information, based upon its supporting evidence. Some systems use this feature to rank multiple conflicting opinions for display of most likely information first. It is not intended to eliminate the receiver's need to evaluate the evidence for themselves. 

0 =Unreliable evidence or estimated data 
1 =Questionable reliability of evidence (interviews, census, oral genealogies, or potential for bias for example, an autobiography) 
2 =Secondary evidence, data officially recorded sometime after event 
3 =Direct and primary evidence used, or by dominance of the evidence 

 
There is some discussion here where levels are expressed as certainly, probably, possibly and unlikely. This is more meaning full than numbers or the Gedcom examples. 

So I suggest internally we also have levels 0 - 3, but in order to express relative priority between choices allow an optional decimal qualifier as e.g. 1.1, 1.2 etc.  Externally the level, if present, should be shown as  certainly, probably, possibly or unlikely.


'lego audio video erro ergo disco' or "I read, I listen, I watch, I make mistakes, therefore I learn"
Posted Wednesday, July 22, 2015 - Post #35265
Famous Writer

Famous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous Writer

Customers
FamilyTrees.GenoPro.com
GenoPro version: 3.0.0.4

Last Login: Sunday, April 02, 2017
Posts: 319, Visits: 2,117
Is this not "a bridge too far"?  What good would it be to know that a person actually died on May 1, 1680 when everybody believes it was April 29?

Looking at the certainty assessment, level 3 would only be valid if you have witnessed the event yourself, given you were sober at that time.

Based on the errors I found in official registers, all official public records would receive a 2 from me.  I've seen children registered under one given name and baptised under another.  I have an grandmother buried in France under the name of her daughter.  I've seen 6 out of 8 children in the same household registered under the wrong mother (a niece of the biological mother with another first name), etc.  One thing is sure, I haven't seen it all.

Wouldn't the simplest solution be to put an * behind the item concerned which tells you to read the small print, in casu the comments?

I know I would not be obliged to fill in that type of information and that I'm free to use the comments fields if needed.  But again, how certain can you be that the score you give is right?  I'm a little bit concerned about the effort needed for such an extra development.  That's all.
Posted Wednesday, July 22, 2015 - Post #35266
Legendary Master

Legendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary Master

Administrators
Moderators
Customers
Gamma
Translator
GenoPro version: 3.0.0.7

Last Login: 2 days ago @ 12:03 PM
Posts: 4,365, Visits: 21,495
I like the concept of: certainly, probably, possibly or unlikely.  Of course, there would be the 'blank' option which would be by default, meaning the user did not assess the credibility.  I think the blank field would  be the value 99.9% of the time, however in some situation, the user would pick between certainly, probably, possibly or unlikely when there is inconsistent/incoherent data.

Another feature I am considering of having the following for every object (individual, family, picture, birth event, death event, etc):
  • Note.  This is what is current the 'comment' field, however I think the term 'note' is more accurate than 'comment'.
  • Personal Comment.  A personal comment is some extra information related to research, including a 'todo' list.  By default, the personal comments would not be exported into the JSON, so they would not show up in the report.
  • Discussion.  Discussions are personal comments from multiple users, where each comment recorded with its author/writer and date & time, like a group chat conversation.  GenoProX already has a built-in chat, and it would be possible to add a discussion to every object.  The purpose of the discussion is not to gossip about people, but provide a tool for people collaborating on a family tree and having a richer mechanism to document their conclusion on incoherent data.  Say for instance, there are two documents with a different date of birth, such as a baptism and census record.  Well, a discussion would explain why the baptism date was the correct date because the census record was filled by the parents who had 14 children and swapped the month of birth with another child.  Like the personal comments, by the default GenoPro would not export them to a report.
As far as storage and performance is concerned, this feature would not require any extra memory if the note, personal comment or discussion are empty.  Despite all the new features in GenoProX, the new architecture of GenoProX is likely to take slightly less memory (RAM) for opening the equivalent document created with GenoPro 2016.  Of course, since GenoProX will save the 'undo history' in the document, the document will require more memory as the undo history grows (there will be the option to flush the undo history).  The speed of serializing (saving) into JSON and unserializing (loading) from a JSON document is likely to increase by a 10 times, perhaps 100 times the performance of GenoPro 2016.

Edited: Wednesday, July 22, 2015 by GenoProSupport
Posted Wednesday, July 22, 2015 - Post #35267
Junior Member

Junior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior MemberJunior Member


Customers
GenoPro version: 3.0.0.9

Last Login: Monday, January 23, 2017
Posts: 12, Visits: 29
I really will love to have such status.
I'm French, in my region, until around 1910-20 a lot of people not used they firstname, but another one must often the second or the third one, you have always to check because only birth and death give you the must accurate firstnames !
I have one case Marie-Thérèse Verry, her birth, death, marriage are with Marie Thérese, one of her child with Marie, 3 with Catherine, when her husband died he was married with
Marie-Catherine ! On Census, each time another firstname...
I have also some of children who died in an early age and the order of the firstnames has been switch
meanwhile !

I really would like to have a "uncertain" status, this will help me to identify quikly (or after a long time without working on my tree) where stranges things are... Because the Comment field begin to become a "
hodgepodge"
Posted Wednesday, July 22, 2015 - Post #35268
Legendary Master

Legendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary MasterLegendary Master

Administrators
Moderators
Customers
Gamma
Translator
GenoPro version: 3.0.0.7

Last Login: 2 days ago @ 12:03 PM
Posts: 4,365, Visits: 21,495
What about the following list:

certainly, // 99% +2
credibly, // 90% +1
probably, // 50% +0
possibly, // 10% -1
uncertain, // 5% -2
unlikely, // 1% -3
error, // 0% -4 // This information is confirmed to be incorrect, thus an error

The numbers from 0% to 99% could be a way to compute a 'score' of accuracy.  Another method could be to subtract to the score when the information is inaccurate.  Of course, for the user, he/she would pick between certainly, credibly probably, possibly, uncertain, unlikely, error, however internally GenoPro could do some arithmetic to better determine the quality of information.

Edited: Wednesday, July 22, 2015 by GenoProSupport
Posted Wednesday, July 22, 2015 - Post #35269
Famous Writer

Famous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous Writer

Customers
FamilyTrees.GenoPro.com
GenoPro version: 3.0.0.4

Last Login: Sunday, April 02, 2017
Posts: 319, Visits: 2,117
In the meanwhile you could create and add a few "custom tags" to help you make notes.



In table view you can then easily "identify strange things"

Posted Tuesday, September 29, 2015 - Post #35821
Famous Writer

Famous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous WriterFamous Writer

Customers
Gamma
GenoPro version: 3.0.0.7

Last Login: Wednesday, August 16, 2017
Posts: 227, Visits: 751
genome (21-Jul-2015)
I think it can be said that no family tree will ever contain the whole truth, but it should as far as possible contain all discovered facts about the ancestors and ideally the sources of those facts. 

As appleshaw and others have pointed out, the facts can vary depending on the source of the information, memories fade, errors creep in on transcription, different censuses record the same family member differently, people may deliberated obscure their past. etc.

Therefore we should be able to record all variations in the data discovered.  I suggest the following for GenoProX:

  1. Individuals can have multiple names, not just the 'Alternative Name' available currently but multiple sets for all name fields (forenames, last names, nicknames etc).
    As well as coping with the variations in sources this will also help for example in cases where immigrants from places like Russia and Poland to the UK changed their names to 'Anglicised' versions, also where people have different names in other languages e.g. Hebrew and English.

This would be particularly beneficial to me, as I'm working extensively with mediaeval genealogies in which many people are known by English, French, and German versions of their names. The question is, where do you draw the line? Two languages? Three? I can see dealing with more than three variations on an individual's name becoming unwieldy in the interface, but if the program user is putting the data in, then it should be possible to do so in the most effective manner possible, and adding alternate names to the Notes field doesn't allow proper report indexing. Just a few more thoughts to consider.

GenoPro: Best. Genealogy. Software. Ever.

Edited: Tuesday, September 29, 2015 by GenoProSupport


Similar Topics

Click to view RSS...
Expand / Collapse

Reading This Topic

Expand / Collapse