Thursday, May 10, 2012

I've Become One Of Them...

Be afraid. Be very afraid.

Or laugh if that suits you better. ;-)

After indexing thousands of records for FamilySearch, I took the plunge. Yes, I've become a FamilySearch arbitrator. As of today I've arbitrated just over a thousand names. That's not a lot compared to those who have been arbitrating for years or even months. But I feel like I'm out of the "newbie" category now. I thought I'd share a few of my thoughts and observations on the 1940 Census project since that's entirely what my indexing and arbitrating efforts are focused on at this point. 

I suspect there will be more "errors" in the finished product (the index for the 1940 Census that will appear on, and than for previous Census years. The reason for this is the poor image quality we indexers are working with. In addition to having to decipher a zillion different enumerators' handwriting, as is always the case with census records, we also have to try to read unfocused camera images. As someone who never leaves the house without a camera, I'm very attuned to focus issues. When handwriting is out of focus it creates ghosting. It's almost like having double vision. It complicates deciphering even the most precise printing and can make sloppy handwriting beyond difficult to decipher. I find deciphering ghosted handwriting much more difficult than deciphering sloppy handwriting or handwriting using "old" letter forms. There are samples galore of "old" handwriting styles to reference but I've not seen any for deciphering ghosted images. Fortunately, not all the pages are out of focus. But I would go so far as to say that most pages have some soft or blurred sections and some pages have focus issues throughout. If you've tried to look for family members among the 1940 Census images posted online you know what I mean. 

Fortunately, the folks at FamilySearch have a procedure for dealing with indexing errors. When you find one (after the indexes are published) you will be able to send them a correct/alternative spelling. I suspect there will be a lot of that going on. Hey, we're all doing the best that we can but you can't make a silk purse out of a sow's ear. 

I erroneously assumed that the most common errors I would be correcting as an arbitrator would be differences of opinion on letters within a name/location. But my experience has been that the most common errors I'm correcting are typos. Sure, sure I'm doing my share of correcting spelling but I'm doing way more correction that has to do with keyboard entry errors. 

It's nice and productive to get a rhythm going when you're entering data but sometimes that "efficiency" comes at the cost of accuracy. I'm finding that folks are often missing column changes due to the auto-fill feature. Auto-fill is really handy. It will fill a given cell with the content from the previous time you typed a word starting with a given letter in that same column saving time and typing. That's a good thing. Sometimes. But often people don't catch that the last time they typed the letter "S" to start the relationship column it was for "sister" but this time they meant "son". They type the letter "S" and hit the tab to the next column assuming the auto-fill filled in "Son" like they've typed several times on that page already. But! The last time they typed it, it was for "sister" and that's what the auto-fill remembers. It's an easy mistake to make when you've got a rhythm going. 

Probably the next most common error has to do with the city location for where people were living in 1935. Specifically, the letter "R" is the issue. "R" is an abbreviation for "Rural". The instructions to the indexers for this column of information reads as follows... (emphasis is mine)
The entry in column 17 may be a place-name, or the census taker may have written 'Same house,' 'Same place,' or 'Rural' or some abbreviation of those terms, such as 'R,' 'SH,' or 'Same H,' in this column or across multiple columns.
If a place-name was entered in this column, then index the place-name in this field. If the place-name was misspelled, spell it correctly. If it was abbreviated and you can determine what the abbreviation stands for, type the complete name instead of the abbreviation. Use the lookup list for assistance. If you are not sure what the abbreviation stands for, index what was written, excluding punctuation.
If 'Same house' or 'Same place' or some abbreviation of those terms, such as 'SH' or 'Same H,' was entered in any of the three residence columns for April 1, 1935, then index the term, as it was written, in this field.
If 'Rural' or 'R' was entered in this column, then index the term, as it was written, in this field.
It's so easy to overlook that last instruction! When it comes to an actual place-name (name of a city, village, town, etc.) you are supposed to correct spelling and type the complete name instead of the abbreviation. However, when it comes to the R/Rural designation you are to "index the term, as it was written". Which means, if you see an "R" you type an "R". You do not write out the complete word "Rural" instead of the abbreviation. I've had arbitrators miss this too and erroneously change "R" to "Rural" on several pages that I have indexed. It's frustrating and it brings your accuracy score down. And you know that's not what the instructions say to do. And you so want to send a little note along to the arbitrator to point out their misunderstanding, but you can't.

What can you do? Chin up. Carry on. Keep the big picture in mind and don't sweat the small stuff. Don't let the letter "R" issue keep you up at night, lol!

My Part
I'm a part of the Michigan Genealogical Council group, the third most productive group of indexers according to the latest statistics. At this point, I am indexing and arbitrating the 1940 Census for the state of Michigan only. I've indexed census sheets from 21 counties so far, Alcona, Allegan, Arenac, Baraga, Barry, Bay, Berrien, Branch, Calhoun, Cass, Charlevoix, Chippewa, Clinton, Delta, Dickinson, Eaton, Emmet, Genesee, Gladwin, Gogebic, and Grand Traverse. More counties have been done but I skipped indexing for several days last week when I was out of town. There are a total of 83 counties in Michigan.

I've come across a good number of Polish surnames with all their many consonants. It must be a nightmare for those indexers and arbitrators who aren't familiar with them. And for the enumerators too for that matter. For me, when I see them, it brings familiarity. Not that I haven't come across some unique Polish surnames, I have. But most of the names I've some across have common letter combinations.

The indexing for Michigan has been moving along at a good pace. However, we haven't gotten to the really densely populated counties yet (Wayne/Detroit, Oakland and Macomb/Detroit suburbs, Ingham/Lansing, and Kent/Grand Rapids). I'm sure the rate at which we move through the counties will slow down when we do.

It feels good to be contributing to the cause and knowing that I will be helping others find their family members in the 1940 Census. Have you thought about helping out with 1940 Census indexing? It's not too late! We can always use more help. You can download the software to get started here.