How do I check to see what my percentage of correctness is for my indexed batches

DixieMcphie1

How do I check to see what my percentage of correctness is for my indexed batches

erutherford

Feedback and batch correctness are not features in web indexing.

DixieMcphie1

we used to be able to see it, why not now

Hiram Bertoch

There were hard feelings being generated. Even very experienced indexers took some of the corrections personally. It was more peaceful to the community not to have people be able to see how they had been correct.

barbaragailsmith1

Most of the bad feelings came from indexers seeing how arbitrators changed their correct entries. When the feature where we could see our errors first came out, Family Search was inundated with calls from indexers about their correct work being changed. The folks at Family Search were as stunned as everyone else at how many arbitrators didn't follow the indexing instructions. It was a mess and it was decided that it might be best to take that feature away. Which is sad for the indexers whose work was being correctly changed by competent arbitrators because it was a good learning tool for them.

DixieMcphie1

Thank you, I guess that makes sense because I know how hard they work, and even we arbitrators have to learn. Thank you for being there for us, so we can learn from our mistakes. Dixie McPhie

John David Durham

Perhaps a better approach would be to create a pre-test before indexing/reviewing. Currently, there is a message asking indexers/reviewers to read all instructions before proceeding to work. In my opinion, this is rarely done by a indexer or reviewer. Having reviewed close to 750,000 records over the past few years, I notice the errors that are done are mostly because the indexer/reviewer has not read the instructions. The current City Directories project is a prime example. I've reviewed over 50,000 of these since Feb 1, 2024 and have noted these mistakes as most common:

1. Including the widow when indexing — Smith, Sally Jo wid John Smith
   1. The instructions say NOT to include John Smith as the spouse in this case. This should be indexed as Smith, Sally and omit John Smith as the spouse.
2. Omitting the surname when listed as John Smith (Sally Jo)
   1. The instructions say to include Sally Jo as the spouse and to assume the surname is Smith if not stated otherwise
3. Including Jr as a prefix or part of the indexed given name
   1. The instructions make it clear that JR is not a prefix and should not be listed in the Given name field as well. It should be omitted.
4. 1. Including multiple names using Or as a separator. Smith see also Smyth . I see many instances of indexing the surname as Smith OR Smyth. Not appropriate when the phone index is merely suggesting that if you don't find the person you are looking for here to look elsewhere

These seemingly minor issues seem trivial, however, they take time to review and correct. There are other issues as well.

My suggestion would be to have all read the instructions, then attempt to correctly index the test batch. This batch should include most of the issues that regularly occur. Once the batch is completed, the batch could be scored by computer and returned to the indexer with the problems that occurred with notes about why they were noted. The test would not be seen by anyone other than the indexer giving them instant feedback with no hard feelings.

This process would no signifantly slowdown the process and would provide the indexer with a good foundation before proceeding.

With that said, it could cause a bit more work on the front end for the batch creators/Instruction writers to carefully choose a batch that contains plenty of opportunities to get a few issues incorrect unless the indexer read the instructions or could provide the indexer clarification about an issue they were unsure of and didn't want to ask.

Thank you for allowing me to share my opinion.