Where do I find accuracy of indexing projects?
Where do I find how I am doing on indexing or reviewing?
Best Answer
-
Thank you all for answering the question re: accuracy! YAY.
1
Answers
-
There are no reports provided to us on our accuracy rates.
0 -
I heard from some of the long time indexers here that family search used to do an accuracy report and sent it back to each user, but they have since discontinued that. Now it just records totals. Id say if you are doing indexing, double check your info before you submit, so that when a reviewer checks, then double checks your work, its really a triple check. Its the trust we have while we do our best that everyone else is doing their best too. Hope this helps. :)
3 -
FamilySearch didn't really "do" an accuracy report. The old system had two indexers index the same batch. If the indexers were in total agreement, the batch went on to publication. If they were not in agreement, then the batch went to arbitration (reviewing). Many of us arbiters did receive training or had indexed significant numbers of records (like 30,000) to receive the right to review the batches. What the indexer received back were the results of the arbitration and a percentage of agreement with what they had indexed. The problem was that if the arbiters didn't read the project instructions and didn't look at the field helps they made lots of mistakes. (The same mistakes that will happen if the reviewers don't do their due diligence on every project they review in web-indexing). This caused a lot of problems for indexers and was quite confusing.
If you are reading the project instructions, reviewing the examples, and reading the field helps, and have read all of the General Indexing Guidelines at the bottom of the project instructions, you are probably doing a fine job. If you are unsure of an instruction or need help with a batch, someone will be able to help out on this community page. Just the fact that you are wondering about your accuracy for me is an indicator that you probably are spot on with your indexing.
4 -
Check out this link to a suggestion to bring back some form of direct feedback between Reviewers and Indexers. You may decide that you want to support (up-vote) it. I made some comments there that pertain to this discussion.
I agree with Melissa that the key to becoming an excellent Indexer and subsequently, an excellent Reviewer is to read, understand, absorb and diligently apply all the Project Instructions, Field Indexing Examples and General Indexing Guidelines that come with each project. And to reread them when you start a new part of an old project. Instructions can change over time as new parts are released.
But, I also believe that there is an important role for direct, real-time (anonymous) feedback from more experienced to less experienced Indexers/Reviewers. And perhaps feedback in the other direction too. Even experienced, earnest Reviewers make mistakes sometimes. As Melissa has pointed out, not all Arbitrators were equally skilled or responsible and the poor ones caused many to get upset, including myself and Melissa.
Because of the consternation with the old system, along with the move to Web Indexing, we lost all forms of direct feedback. BTW - we also had mentors in those days. At the same time, the standard for being allowed to check the work of the Initial Indexer (I.e.becoming a Reviewer/“Arbitrator”) was dramatically lowered from Indexing many thousands of records with a high level of “accuracy” to simply Indexing 1000 records, with no idea of how well they are done.
I have recommended that Family Search at least “Pilot-Test” an opt-in on both sides procedure in which willing Indexers could receive feedback on their work from willing Reviewers - who might be willing to receive feedback too. I think the current Help>Labs optional Reindex Batch with “Reasons for Reindexing” form is a good starting point as a feedback mechanism. However it inserts a FS support person between the Reviewer and an evidently very ill-informed or even malicious Indexer. I recommend direct feedback at an earlier stage, hopefully to forestall such egregious errors that require reindexing. It wouldn’t stop malicious behavior, however, so there might still be room for that form and process.
3
