Indexing and reviewing advanced records
I notice so many errors when reviewing advanced records, that I hesitate to refer this website to any serious genealogy searcher. This also leads me to believe that correctly-indexed advanced records are being ruined when otherwise correct entries are undone, resulting in a database that makes it impossible to expect accurate results when searching the FamilySearch records. It appears obvious that some indexers/reviewers are not reading (or are ignoring) the batch instructions.
My suggestion is this: Prior to an indexer/reviewer being able to open an advanced record, there should be a pre-test. If the indexer/reviewer passes the pre-test, then the record to be indexed/reviewed will open. The test could include some simple true/false questions that highlight common errors, like using punctuation marks in names, expanding spelling entries where they should not be expanded or corrected (including names like "Thos", entering a child's surname when the surname should be assumed, using a comma to separate place name levels, etc. Indexers/reviewers should not be able to circumvent the test page.
In each case, the correct answer could be revealed. This would have the effect of forcing an indexer/reviewer to read the batch instructions, thus assuring more accurate FamilySearch records.
Comments
-
Indexes are "finding aids" - not the record.
If a person indexes Thomas or Tho*s for Thos (what displays in the record) - is that going to harm the researcher from finding the record and making their own conclusions about the record? Is the name really Thormopolous?
I have found many records indexed incorrectly - it didn't make me think less of FamilySearch at all. And if editable I could correct or if not submit that to FamilySearch. Some have been corrected - there are more to be corrected I'm sure. FamilySearch has been working on that solution.
The key to indexing is - "do your best" - not "expect perfection". My best might be different than someone else's - and the Reviewer can do their best and change what I index... "All hands on deck" - "the more the merrier"... AI indexing ... hmmm ... I guess it's ok ... As long as reviewed before publishing?
I personally would not be opposed to a test taking me through the Project Instructions - but "my best" still might be different than others.
2 -
What such a test would achieve is unindexed records.
Yes, indexers fail to follow the instructions quite frequently. But as genthusiast said, is the published index really going to suffer if Thos. is expanded to Thomas, or a child's surname is assumed based on the parents? Heck, my all-time favorite misindexing skipped the father, changed the mother into the father, and created a mother out of the father's occupation and religion. But the moral of the story is, I found the record. Anything really is better than nothing.
2 -
After reading the thread at https://community.familysearch.org/en/discussion/121510/rhodesia-voter-registration-project#latest I came to the conclusion that following project instructions can be the least useful action for an indexer, if they want their work to produce useful results for researchers. But (amazingly to me), after all his criticisms, the original poster goes on to say: "At the end of the day PI needs to be followed no matter how flawed it is." Sorry, but I'll never understand the reasoning behind such illogical advice, which most indexers seem to find perfectly acceptable. There should be far more oversight of project leaders, to ensure their instructions are not plain crazy (as in indexing a placename known to be incorrect, in this example).
The PIs for some projects appear to be so flawed that the intended end result - of providing accurately indexed records that are of genuine use to researchers - is proving be a failure.
1 -
We made a note of your idea/comment and have forwarded it to the engineers.
0
