*RELIABILITY settings

Questions regarding use of any Version of Family Historian. Please ensure you have set your Version of Family Historian in your Profile
Post Reply
jjensen
Newbie
Posts: 4
Joined: 01 Jan 2019 15:27
Family Historian: V6.2
Location: Denmark

RELIABILITY settings

Post by jjensen » 01 Jan 2019 15:33

Hello

I have to set RELIABILITY for many of my data. fx RELIABILITY (2 _PROOF proven), so I and others are sure we have papers on that data.
Where can i place "2 _PROOF proven" so it's saved together with the data???

Thanks in advance
Jjensen

User avatar
tatewise
Megastar
Posts: 16176
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: RELIABILITY settings

Post by tatewise » 01 Jan 2019 15:57

Welcome to the FHUG.

I guess you are new to FH and was able to use a RELIABILITY setting with a 2 _PROOF tag in some other product?

That is a non-standard GEDCOM tag (it starts with underscore) that was unique to that other product, and is not supported by FH.

The standard GEDCOM has the ASSESSMENT (QUAY tag) associated with Citations of Source data.
See Knowledge Base > Information recorded within a citation.
Does that make sense, or do you need more explanation?
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 01 Jan 2019 16:02

Are you looking for a way of determining that something is completely unsourced (ie. there are no source citations to support the fact) or that there is sufficient supporting evidence to have a high degree of confidence that the fact is proven?

Mike's mentioned the citation assessment which tells you a hybrid mixture of detail about the source but doesn't indicate how robustly the overall fact is proven.

If we know which it is we might be able to suggest ways forward.

(You may already be familiar with it but the Evidence Analysis Process Map may be of interest -- FH does not implement this, unfortunately.)

jjensen
Newbie
Posts: 4
Joined: 01 Jan 2019 15:27
Family Historian: V6.2
Location: Denmark

Re: RELIABILITY settings

Post by jjensen » 01 Jan 2019 16:14

Hi again.
No i'm not new to FH, but my problem started when I should give a copy to the family. For that I used Gigatrees, very nice prg, and found that I was missing RELIABILITY settings.
I'm living in Denmark and we have nearly all church-books online.
Therefore it will be a lot easier for other in the family to look for what is missing if I can set RELIABILITY settings in FH.
Thanks in advance
Jjensen

User avatar
tatewise
Megastar
Posts: 16176
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: RELIABILITY settings

Post by tatewise » 01 Jan 2019 16:38

OK, sorry, because you just joined today I thought you were new to FH.

As Helen suggests, you need to give more detailed examples of what data you want to give RELIABILITY values.
Are you talking about Facts (Events & Attributes), or Citations, or Source records, or what?
These should be familiar concepts if you are familiar with FH.

It seems that http://gigatrees.com/blog/gigatrees has a Reliability Assessment feature defined as follows:
When publishing a family tree it is useful for visitors to be able to quickly and accurately assess the validity of the claims being made. Gigatrees automatically calculates the reliability assessment for every claim using several pieces of data including, the source category defined for the referenced sources, the source reference quality (QUAY) defined within the source references for the claim, and the source type defined within the source itself. Gigatrees will display the calculated reliability assessment alongside each claim. Gigatrees also supports manually setting the reliability assessment for claims.
The "source reference quality (QUAY)" is the Citation Assessment (QUAY) I mentioned earlier, so if you set that it should be reflected in Gigatrees. That is the only standard GEDCOM tag available in FH. Whereas _PROOF and _QUAL are non-standard GEDCOM tags only supported within Gigatrees
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 01 Jan 2019 16:50

Mike,

The Gigatree reliability assessment relies not just on Source Quality (QUAY) but also a 2 source categories associate with the source (custom fields); it can be overridden by assigning a _PROOF tag to the fact. It sounds as if Jjensen wants to set that _PROOF value to do all the assessment manually.

Even setting Source Quality wont help as it's only a very minor part of the Gigatrees calculation.

I'm at a loss on how to proceed -- how to set a 2 _PROOF tag without editing the Gedcom directly.

jjensen
Newbie
Posts: 4
Joined: 01 Jan 2019 15:27
Family Historian: V6.2
Location: Denmark

Re: RELIABILITY settings

Post by jjensen » 01 Jan 2019 17:08

Thanks for the help with asking in the right way. Yes i can manuelly do it in the gedkom file (it have more than 60 000 lines), but when i edit next time i have to do all the work again. (not so nice)

JJensen

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 01 Jan 2019 17:14

Yes, manual editing would not realistic. I'm hoping Mike might have a good idea -- maybe a bit like Sort Dates and handle them on the export with special Gigatrees output Mike? (Don't worry ir you don't understand this Jjensen -- all will become clearer if it's a viable way forward).

User avatar
tatewise
Megastar
Posts: 16176
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: RELIABILITY settings

Post by tatewise » 01 Jan 2019 18:00

Helen, yes, my first reaction is to suggest a labelled Note meta-field per Fact such as:
[[Proof: Value]]
Then the Export Gedcom File Plugin, with a new Gigatrees option, could create 2 _PROOF Value tags.
BUT it still requires a manual edit for each Fact to add those labelled Note entries.

Gigatrees offers several ways of using existing features in FH to set Reliability, so why not use them?

You say "Even setting Source Quality wont help as it's only a very minor part of the Gigatrees calculation".
I don't understand. If the only parameter supplied is QUAL tags, then that is what Gigatrees will use, and little else.

JJensen, the http://gigatrees.com/blog/certainty-assessments blog explains how the assessment is calculated.
"When multiple sources are referenced for a claim, the source with the best quality is used."
Later in the table it says that the Citation Assessment QUAY tag is handled as below:

Code: Select all

 QUAY   Assessment   Gigatrees Reliability
  0          Unreliable        questionable
  1          Questionable   proposed
  2          Secondary       probable
  3          Primary            certain
So, the Citation Assessment with the highest value for any one Fact will set the Reliability in Gigatrees.
e.g. For a Fact in FH set one Citation Assessment to Primary then the Reliability of the Fact will be certain in Gigatrees.

Does that provide you with what you need?

If not, then please explain why it is not a suitable solution.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 01 Jan 2019 18:18

Mike, relying only on QUAY misses out a whole chunk of the Evidence Analysis process (all the Source Category stuff). You might want to look at the Evidence Assessment Process map that I referenced upstream.

Manual editing of the fact note (and ideally the source note to add the Source Category stuff) would only need to be done once per fact (unless extra evidence emerged, when you would need to change it) and once per citation (when creating it).

User avatar
tatewise
Megastar
Posts: 16176
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: RELIABILITY settings

Post by tatewise » 01 Jan 2019 18:58

I may have misunderstood, but I thought "When there is no other information available to determine the reliability assessment, Gigatrees makes one last ditch effort by looking at the source's type (@SOUR.TYPE or @SOUR._TYPE) field."
But if source categories are not enabled, then they play no part. Isn't that the default?
I was going to introduce those options, but did not think they would be necessary.

Anyway, my interpretation is that JJensen only needs a simple Certain v Questionable reliability assessment.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

jjensen
Newbie
Posts: 4
Joined: 01 Jan 2019 15:27
Family Historian: V6.2
Location: Denmark

Re: RELIABILITY settings

Post by jjensen » 01 Jan 2019 19:08

Hi again
I have now tried to edit the original gedkom file. It look like it's possible anf FH will accept to run the file again.
I have added "2 _PROOF proven" after a couple data, saved the file and opened again in FH. Export to Gigatrees and it looks OK.
Don't hope the file will be damaged from the editing!!

JJensen

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 01 Jan 2019 19:13

Yes, but it's a very poor fallback, Mike -- it sacrifices all the sophistication of the Analysis Process. QUAL only allows you (for example) to mark something as Primary or Secondary, not Original or Transcript or Copy or Derivative (with all the possibilities for error that moving further away form the Original introduce) -- they're different axes of assessment. A source that is Primary but consulted in Transcript (think about the old IGI) is less reliable and possible less complete than a Source that is Primary and consulted in the Original (an original Parish Register).

If this were implemented fully, I'd utilise it as well -- it's the missing piece that I used to have in GenQuiry and have sacrificed. Now that I know Gigatrees supports it, I might take a look at Gigatrees myself.

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 01 Jan 2019 19:15

Jjensen, it should (if what Mike and I are discussing can be implemented) be possible to allow you to do the edit within FH (by adding specific text in the Note associated with the fact) and then export it to Gigatrees in the format Gigatrees expects (using the Export Gedcom plugin) -- that would eliminate any of the risk associated with direct Gedcom editing.

User avatar
tatewise
Megastar
Posts: 16176
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: RELIABILITY settings

Post by tatewise » 01 Jan 2019 19:55

JJensen, as Helen says, editing the GEDCOM file is risky and typos are a problem.
Also, FH treats _PROOF tags as UDF, see Knowledge Base > Uncategorised Data Fields.
That will make handling genuine UDF a bit more difficult.
When exporting the GEDCOM to any other products, the non-standard _PROOF tags might cause problems.

Please try the Citation Assessment technique that avoids editing the GEDCOM and needs no Plugin development.
It will provide what you need if you only want to differentiate between the four values questionable, proposed, probable & certain.

Helen, yes, I accept it sacrifices sophistication but JJensen wants to use _PROOF tags that also override that sophistication.
A snag with the labelled Note idea is there is no syntax check on the Proof: values, so typos are a problem.
Using Citation Assessment QUAY tags is much the same, but avoids typos and avoids editing the GEDCOM.
Can we start a new Plugin Discussions thread for any further technical investigations.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 01 Jan 2019 20:01

Mike, happy to start a new plugin discussion, but I would like to make the point that using the Citation Assessments will deliver unpredictable results in some cases -- if there are multiple sources cited, working out how to "massage" each one to get the result wanted is bad practice as well as being unreliable.

Edited to add: Jjensen, are you using Gigatrees Pro -- no longer available to new users -- or Gigatrees Mini which seems less sophisticated in this area?

User avatar
tatewise
Megastar
Posts: 16176
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: RELIABILITY settings

Post by tatewise » 01 Jan 2019 20:31

What about the rule "When multiple sources are referenced for a claim, the source with the best quality is used."
Doesn't that deliver predictable results if only QUAY is used, and all other features aren't used or left in default/disabled mode?
I assumed it was simply a choice of the highest QUAY value.
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 01 Jan 2019 20:43

Mike, QUAY (Assessment) and _QUAL are different things.

If you only use QUAY, you are basing your assessment on whether the information you're relying on in the source is primary or secondary or something that you're not sure about at all... it says nothing about whether you worked from a dodgy photocopy of a handwritten transcript done on the vicar's study floor by a bored fourteen year old boy (sitting on my bookcase a meter away -- and even he doesn't consult it any more because there are better alternatives) , or a digital image of the original register cleaned up and put on line. They're both technically Primary Information (using the QUAY field) but the second one is more reliable as it's closer to the original and less likely (but not impossible) to have lost data in transmission (_QUAL).

Yes, you could manipulate QUAY to drive the outcome you want, but you'd be misusing them because (in the case above) you'd have to lie and say that one of the sources is NOT a representation of Primary Information when it really IS.

This may be moot, pending Jjensen's answer to which product he is using.

User avatar
tatewise
Megastar
Posts: 16176
Joined: 25 May 2010 11:00
Family Historian: V6.2
Location: Torbay, Devon, UK
Contact:

Re: RELIABILITY settings

Post by tatewise » 01 Jan 2019 23:17

OK, point taken, but as you say probably moot, since JJensen specifically mentioned Danish church-books online.

BUT adding a [[Proof: value]] override _PROOF feature does nothing for setting Quality _QUAL, so maybe the Plugin Discussion must cover all those features?
Mike Tate ~ researching the Tate and Scott family history ~ tatewise ancestry

User avatar
ColeValleyGirl
Megastar
Posts: 1192
Joined: 28 Dec 2005 22:02
Family Historian: V6.2
Location: Cirencester, Gloucestershire
Contact:

Re: RELIABILITY settings

Post by ColeValleyGirl » 02 Jan 2019 07:19

I was taking the Danish Online church books to be an example of a more general application. But yes, the plugin discussion should cover the gamut of aspects.

Post Reply