Sources Transcription or Display BUG?
Please review the attachment and advise if I'm reading this wrong, but the source reviewed below appears to be contradictory in relationships transcribed and/or displayed; or at least VERY confusing!
Besides this issue, highlighted below, this is the most difficult part of the proposed new display of sources that appears to be a part of the Family Tree, and I am having a very difficult time swallowing this pill! Is this truly what is being proposed for displaying sources going forward?
Answers
-
The relationships shown in the table below the current person of record are in terms of that person of record, in this case the elder Thomas Shaw (https://www.familysearch.org/ark:/61903/1:1:SGFK-YKW). The issue is complicated by misindexing of the relationships.
It's probably easiest to see what was indexed by looking at the head of household's entry (https://www.familysearch.org/ark:/61903/1:1:SGFK-YK7). It shows Mary Shaw, age 73, as head of a household also containing Thomas, 47; Sarah, 41; Samuel, 10; Thomas, 8; and Ann, 5. All of the other people were indexed as "son" or "daughter", which is probably what the census actually says -- except it probably indicates different parents for the grandkids. There was no field in the indexing for the "to" person of the relationships, because it assumed that this would always be the person designated as head of household -- which is probably what the instructions for the census called for; it's just not what humans naturally do.
So what the table below the elder Thomas Shaw's indexed record claims is that Mary is his mother, and Sarah, Samuel, the other Thomas, and Ann are all his siblings. The impossibility of this is not a fault of the display, but of the indexing.
3 -
Yes, that is the bug, and in my opinion is a process that should not be allowed: "The relationships shown in the table below the current person of record are in terms of that person of record".
The current person of record should be displayed in the ROLE originally shown on the census record, as it has been heretofore. Having the system generate a new role, i.e. "Mother" in lieu of "Head", "Sister" in lieu of "Daughter", etc. solely based on the current person of record is a distortion of accurate transcription.
Further, "The impossibility of this is not a fault of the display, but of the indexing" is an inaccurate statement, because the system generated "display" is not what will be found on the original record when viewed.
Let the researcher see the limited, accurate data lifted from the original record, based on the indexers best efforts. A seasoned researcher will always consult the original record to assess the information and compare it with the other sources available to interpret where needed.
Indexers are taught to not do exactly what this, apparent system generated program is being allowed to do. I would appreciate hearing from @Sam Sulser or someone associated with this issue to explain why this was allowed, and also respond to the second part of my question (see above).
0 -
You are correct that what is displayed is likely not what will be found on the original record, but this is not the fault of the "system generated 'display'". The fact that both Thomas and Thomas Junior are labeled as sons of Mary Shaw is a fault of the indexing. There exists no tweak of the display algorithm that can correct it.
0 -
The main point here is that the "display algorithm" should be removed from the system, so that researchers can do the work that belongs to them, not to AI. What is not found on the original record should not be contrived by a program, especially when it is not allowed to be done by indexers.
The fault, if there is any, lies with the recording clerk or census taker, or the person who misspoke the information requested by the clerk or census taker. However, to interject a fault into the system, which adds inaccurate information to the indexers work that is not found on the original record, is perpetuating and expounding the error.
The "tweak" is to remove the flawed "display algorithm" design bug, and return to the indexing rule that requires only what is found on the original record - no more and no less than is specified.
0 -
Let me reiterate what you are saying.
You don’t like that web page is showing the relationship from the context of the person you are viewing in the record.
You would like it to always/only show the relationship indicated on the record. In this case the relationship to the head of house.
0 -
@Ottley BQ I hear you! I understand what you are saying about the new display for census records. Don't tell anyone 🤐 but I don't really care for this display either. I think it is confusing. I know for me that part of it is just that I'm so used to it being displayed differently that I have a hard time seeing it this way. I haven't heard the justification but I'll see what I can find out. Maybe we'll both learn something!
Sam 😊
1 -
@lyleblunttoronto1 to reiterate what I am saying, 1) I don't like that the web page is calculating the relationship, implying what cannot be found on the original source record, and thereby violating indexing rules. 2) I would like it to always/only show the ROLE and/or relationship indicated on the original source record, and not overstep the boundaries imposed on indexers of that record.
It appears that the current relationship display algorithm has been added since August, based on the screen prints I doctored up from a previous post (https://community.familysearch.org/en/discussion/129329/source-tab-the-proposed-new-method-is-a-researchers-nightmare), where a suggestion was given of ways to enhance the unpopular method of displaying sources (see screen print below). The additional field added to display a "calculated relationship" has been added (see screen print at the top of this post) to this display during the past month or so, which is the subject of this post.
@Sam Sulser I would be interested in hearing the justification for the unpopular changes being made to the very foundation section (Sources) of every Person Page on the Family Tree. This section does not need change, as much as it needs attention, just like that being given to the Collaborate tab! A true researcher knows it's value, but it's value is being considerably reduced by undesirable UI modification.
1
