Friday, December 17, 2010

Genealogy Software: The Evidence and Conclusion Process debate

This morning we added an introduction to the Evidence and Conclusion Process on the BetterGEDCOM wiki. I'm blogging about it in the hope more will chime in.

Snippets from the developing page on the BetterGedcom wiki:
"Genealogical software was originally developed for the purpose of recording "conclusions" so that when GEDCOM was developed, it's purpose was to facilitate the sharing of conclusion level communication. As the use of genealogical software was expanded, records of "hypothesis," as opposed to conclusions (I "think"/"maybe," vs I know or I have proved) became recorded using otherwise conclusion-level definitions and mechanics built into the software. GEDCOM not only facilitated the sharing of "maybes" as "I knows," but the limitations of GEDCOMs definitions and mechanics no doubt further conflated think/know communication between users. As our use of the Internet expanded, GEDCOM facilitated "publication" of the conflated thinking.
Part of the current BetterGEDCOM effort is directed at ways to untangle this now long conflated "maybe vs know" issue that likely exists, to one degree or another, in genealogical software and is certainly  believed to exist in GEDCOM. Our effort includes developing an understanding of general and scholarly genealogical practices as we begin to define the more specific role software CAN and SHOULD play to  advance best practices within the greater discipline we know and love as genealogy."

If you've read this far, I'm hoping you have thoughts to share!

Personally, I use research logs and prepare research materials/memorandum. I write conflict notes and proof arguments/discussions (often using word processing or spreadsheet software). Most of my thought process, but far from all, becomes summarized in my genealogical software citations. I have a tag called "research notes" and use it to record lengthy information from third parties and/or items I've located that might not be about the person I'm researching. (Russ and I have been blogging here about how and when we enter genealogical data to our software.)

Does my practice meet the standard for Kinship Determination? What about the Genealogical Proof Standard? (Did someone say "reasonably exhaustive?") I'm sure an editor would have quite a go at the record in my software file.

So then, What's needed?

Is the need as simple as adding "level of confidence" [see Mills, Evidence Explained ... (2007), p 19] such as certainly, probably .. apparently ... to the otherwise "conclusion level" software entries?
Should we use a "?" technique, by which we mark particular kinship determinations as unsure, or must we go further?

If we want to go further, then before we access data models, I think we need to document the current, functional "best practices" steps in the research and recording process.

We'd like to learn from you about your practices and your thoughts. Join the dialog here or over at the BetterGEDCOM wiki, "The Evidence and Conclusion Process."

No comments: