Indexing Beginners
We find that many of the beginner indexers make too many mistakes because they do not Read the Instructions either in the Project instructions or in the Field Help. It would be useful to have a tutorial that ensures that each beginner has learned to read the Instructions. Or, that there is a Feedback to the patron that tells them what they need to do to improve. Something to ensure that the indexing gets done with as much accuracy and completeness as possible. Can anything be done?
Comments
-
I don't think this is only beginner indexers. I've seen enough shared batches to know that there are also seasoned reviewers that don't read the project instructions, don't understand them, or simply miss important details in reviewing (commas, misspelled names, errors on place names and dates, for example.) Some mods will say these are non-critical errors because they don't affect a search, but, IMHO, a great reviewer corrects every error.
I think it would be helpful to have indexers and reviewers, experienced or inexperienced, to have to perform a tutorial for every project they work on. It would also be smart if reviewers were given a test before being granted review rights. 1000 entries of given names and surnames does not make anyone a great reviewer (i.e., 1300 entries in a City Directory project - poof, you are a reviewer. Yikes!)
We do have the Reason to Reindex Batch form, but, sometimes a batch doesn't really need a lot of work to constitute reindexing. It would be nice to have something where we could still review the batch and at the same time alert FS personnel that an indexer/reviewer might need a little extra training.
3 -
I feel like new indexers should have to pass a basics test before they are able to index projects. Show them basic general rules (e.g. do not assume gender or surnames, use commas when entering locations, how to enter maiden names, etc) and let them practice with materials similar to what appears in batches. Then test these skills before they can index real projects. I'm reviewing advanced projects and I'm seeing too many basic errors like entering jr/sr with names, guessing gender by given names, no commas used with locations, assuming race and surnames, "correcting" names, typing names in the prefix entry field, and a general TLDR attitude with complex or lengthy documents. No one is perfect but I'm starting to see images where the majority of the entries completed are incorrect.
I also believe that a minimum number of entries completed to access advanced level projects would be helpful. Based on the basic errors I'm seeing and comments where a new user is selecting projects based on interest (i.e. strong interest in Ireland) rather than difficulty level. Advanced projects often require the ability to utilize tools (viewing multiple pages, using ruler tool, rotating pages, modifying brightness) successfully and this takes time and experience.
2 -
We cannot find any "Beginers Index" records to do. Having identified this is the level we want to work on. We are still offered "Intermediate" and "Advanced". Is there a shortage of documents now?
1 -
Attention: @Melissa S Himes & @John Empoliti
0 -
@ClaireBroome you are correct. There are no beginner's indexing to do. You will likely have to go in and find a batch that is fairly easy to do in the Intermediate batches. The project managers are the ones who determine which levels each batch should be rated. There isn't much anyone can do about the quantity of batches offered at a beginner level.
0 -
I also wish there were some way to ensure that Indexers knew and were competent in the primary tools of the Web Indexing Program before tackling real-world projects. We could use training batches with some sort of feedback mechanism. Training videos illustrate many of the tools, and I'm sure that @Melissa S Himes will mention them. I've posted some jpg files (because we can't post pdf files) that describe some of the tools.
The most basic tool that a new Indexer can acquire is the habit of reading (/re-reading), understanding, and absorbing all sections (including especially the General Indexing Guidelines) of the Project Instructions for every batch they Index. Those resources answer at least 80% of the questions asked here.
Regarding the lack of Beginner projects - they do go fast. But, some of the Intermediate projects are attainable for the diligent beginner who practices the habits I mentioned above. For example, the United States—Enlisted and Officer Muster Rolls and Rosters, 1916–1939 [Part O] project comprises almost only typed records, so that big handwriting interpretation hurdle is absent. Some pitfalls surround the Military Date issue, but carefully reading the instructions and asking for help right at the outset can overcome that hurdle.
Many of us would be willing to coach someone indexing a project like that one. This forum is a great place to post a question and get help.
1 -
Beginners projects are hard to come by @ClaireBroome. And I have no idea why so many of the projects have been labelled Intermediate and Advanced, when they don't really need an advanced indexer. They need the same thing all projects need - people who are willing to learn the general indexing guidelines, read the project instructions, look at every example, read the fields helps, and try their best!
This intermediate project that John mentioned is perfect for beginners: United States - Enlisted and Officer Muster Rolls and Rosters, 1916-1939 [Part O]
Another intermediate project which is quite easy is the Canada, Ontario-Tax Assessment Records, 1834-1899
There is an advanced project which if you read cursive is fine for beginners: India - Madras Diocese Protestant Church Records, 1743-1990.
I suggest downloading a few projects and having a look at them. If you think they are too difficult, just click on Return Batch and they go back into the pile for someone else to do.
The video that John referred to is Indexing Discussion (Jason Pierson Live)
I highly recommend it to anyone who wants to participate in web-indexing - even those who indexed on the old program. That link should take you to the video on Youtube. The actual lesson begins at the 8 minute mark. You can find other useful indexing training videos on Youtube by searching with the words FamilySearch Indexing.
2 -
Entering a person's gender when indexing UK documents.
This is a moot point, the instructions say do not enter the gender unless the person's gender is stated. I have reviewed thousands of document entries where a person has not been identified as a "son", "daughter", "wife" or "widow", often the gender box is left blank and equally often an entry has been made in the box.
The problem is that, in British English, a person's given name is indicative of that person's gender, especially so in UK, English documents written up to the early 20th century. So it is not unreasonable for an indexer to make an entry into the gender box based on an english person's name. To lambast people for "guessing" an english person's gender based on their given name is counter productive. Also, while reviewing, I have found many indexed entries where the person's gender has been entered as blank and, yet, their parents' names on the document have been entered as "Joseph & Mary", for example, and the indexed name for the father has been entered as "Joseph" and for the mother has been entered as "Mary", where, surely, the parents' given names should have been entered as "Joseph or Mary", or been left blank.
I would suggest that for UK, English documents entering a person's gender based on the given name is not a problem. If the person's name is unreadable, then enter "Unreadable" in the gender box. Also, for example, if you cannot distinguish between the written "Francis" (M) and "Frances" (F), or "Leslie" (M) and "Lesley" (F), then put "Unreadable" in the entry box.
For UK Catholic, or High Anglican documents which are in latin, the person's gender is always stated -
"filius", son of and "filia", daughter of;
"natus", he was born and "nata", she was born;
"sepultus", he was buried and "sepulta", she was buried.
I understand that indexing is transcribing the document and not translating the document. I can also understand reviewers getting upset by indexers translating abbreviated names to full names or latinized names to english names, but to get OCD about an indexer entering the implied gender of a person is unreasonable.
You have to remember that all of us indexers and reviewers ar unpaid volunteers.
1 -
@PaulHarrison21 You should add your suggestion (copy and paste your post) as a new thread. That way your suggestion could get votes if people like the suggestion that UK records might need an extra instruction that would enable people to use names to determine gender.
I doubt they will change the instructions since we are trained only index the gender listed, or when the relationship status is listed. If the project managers feel that gender can be determined from names (like you suggest in UK records) or occupations, they could make that a special instruction as was done in the Australian and New Zealand records last year.
When we index the names of parents or a groom and a bride, the instruction is that if you can't determine the gender from the names to index the first name in the Father or grooms given and surname fields, and the second name in the mother or bride's fields. (I understand your point, but, Beginning Indexers might not and they might see this thread if they search on Beginner Indexers. We do not index Joseph & Mary in a field.)
1 -
@PaulHarrison21, echoing and expanding on Melissa's reply:
Not indexing the gender based on just the given name is just another specific application of the broad principle that all FS indexing follows: index what you see. If it doesn't say girl or boy, or son or daughter, then don't index "male" or "female". Mark the field blank instead. (Not "unreadable".)
The gender of given names is seldom actually as clear-cut as you seem to believe. When naming their daughters, people have been following the still-current subconscious bias toward boys for ...millennia, now. There is no guarantee that a Jack or Leslie (in any spelling) or Justin is actually male, regardless of the century. Conversely, a Beverly, Shirley, Ashley, or Kimberley in the 1700s and 1800s was very likely male -- but an early bearer of Douglas as a given name was female.
You'll say that you covered these possibilities when you said to put "unreadable" when the gender cannot be distinguished. But what about when the indexer believes unquestionably that Dana must be a woman? Or that Meredith has always been a feminine name? The indexer would be completely wrong in both cases.
Indexing of parents in the appropriate fields is a false parallel: as Melissa points out, parents are identified by position in the entry, not by given name. Fathers are listed first, then mothers. In records where both parents are normally named, if there's only one name, it's the mother (and the child is illegitimate). In records where only one parent is named, it's the father, unless it says specifically otherwise. You don't need to have read the names correctly in order to get them into the correct indexing fields.
3 -
General comment (rehashing of comments): yes indexing is largely 'type what you see into the appropriate fields' - but further it allows you to 'use common sense/do the best you can'. I might agree with indexing position as primary criteria for records with 'column headings' - but for those without perhaps indexer discretion is more applicable.
A relevant example - Irish birth or marriage register listing 'Catherine Murphy' & 'John Sullivan' - Catherine is listed first - so put her in the Father/Groom field?? That's what the GIGs indicate (if there is position rule) ... But obviously common sense will indicate John is father/groom and Catherine is mother/bride. However as an indexer it could be argued that putting position as primary indexing criteria won't change the fact that both are indexed on that one record (having John as the mother/bride though ludicrous will still allow the researcher to find the record and 'laugh at the indexer'...).
Here is an example image from a current Baptism record batch:
Child: John Parent 1: Eliza Dea Parent 2: Wm Whelan
Though the mother is listed first - I certainly am not going to list Eliza in the Father's Given Name Field and Wm in the Mother's...
Though not common I have seen records that change position of father/groom and mother/bride in Irish registers that I have recently indexed. I tend to fall back on the 'common sense/doing my best' clause of the indexer 'contract' when faced with this issue. I assume similar issues occur in records of other countries which may require language fluency to apply 'common sense'...
One complicating factor with Familysearch indexes and this 'issue' (from recollection) is that Source Linker MAY NOT allow you change the position of the index OR if Sex is attached in the index even if you were allowed to change position - there is the additional Sex issue when trying to correctly attach the index to the person in Family Tree (hopefully index Edit options could allow changes but I can also see how it might require a familysearch representative). Perhaps a Familysearch representative can comment on whether researchers reporting these types of record issues CAN get the issue corrected/resolved - and what category/channel is established for resolution of these issues. Community does not seem to be a great conduit for expediting of issues that the user could resolve if Edit were used responsibly (sorry I understand need for data security/integrity).
Reviewer bias: perhaps some reviewers might be overly critical (apply rules over common sense or don't give indexer benefit of doubt)? But they are 'under contract' to do their best ... Good thing >20% correction results in another review pass ... ?
Comment on AI indexing: it appears the AI index may make some records more difficult to Search (names/places horribly wrong). Hopefully researchers will be able to separate the attached from unattached and have a smaller pool from which to manually search/correct.
Conclusion: unless FamilySearch wants to implement more 'oversight' I think variability inherent in crowd-sourced contributions will continue. Theoretically these issues can be resolved when the researcher is trying to attach the index in the Tree (but yes it would be nice to have a more perfect index to begin with).
1 -
Yes, genthusiast, use a bit of common sense. Your conclusion is correct, the indexers are volunteering their time, they are going to make mistakes and those mistakes are what reviewing should correct.
My previous comments about gender based names were referring to UK, Lancashire and Shropshire records. I have indexed and reviewed about 40,000 of these records from the 1640's through to the 1920's and have not found a Shirley, Ashley, Kimberley or Jack as a child's first name. I have seen Beverley as a boy's middle name and Douglas as a girl's middle name, in both those cases they were their mother's surname. I have never seen a Dana in any of the UK records, I think it is a US thing, similarly, Kimberley is an Ausie thing, though it did creep into the UK as a girl's name, usually shortened to Kim in the early to mid 20thC.
1 -
One problem when the review was available to indexers was that some took it personally and some even quit indexing when their 'marks' were low. Much wasted time and energy.
Maybe a better reference for common errors while the project is out. or use Boards (oops indexing board is gone!) 🙄🙁. Or a FB page that would give hints on different projects.
Must be a way. 🤓
0
