I haven’t blogged about the 1940 US Census here for a bit. I actually started this blog post ten days ago and forgot about it. A good blog post from Amy Johnson Crow about arbitrating reminded me of it.
There are some issues with indexing and arbitration of the census. Unfortunately, there are a lot of rules that are being ignored. When indexers don’t read the rules, you hope that the arbitrators fix things. When the arbitrators don’t read the rules, well, the indexers get pissed off because they did read the rules and did things correctly. And their arbitration percentages drop anyway.
Randy Seaver of Genea-Musings has analyzed his arbitration results across several blog posts, finding many of the same things that I have. (Although, they probably aren’t on the first page of his blog anymore.)
I bumped myself up to arbitrator a short while ago. As a group administrator, I learned that I have that superpower — not the power to be an arbitrator, but to make myself one. I think some people might say I should index more, but I was getting so annoyed by bad arbitrators that I just couldn’t. I joined a Facebook group specific to sharing batches and discussing indexing, I have read all the project updates and every page of rules for how to index, and I think I know what I’m doing better than some. And having watched other indexers complain, I know which rules many arbitrators keep getting wrong.
As an arbitrator, I’ve learned even more about indexing and the project rules.
Arbitration is partly about knowing the project rules and following them, and partly about judgement calls. Is that messy squiggle an S or an R or an E? There are three possible opinions and it’s only the arbitrator’s that really counts in the end.
Some arbitrators don’t pay any attention to what they’re doing. Many still leave line numbers on blank lines or change R to Same House when the rules say to index the R. I had one arbitrator that changed age 21 to 1, for the wife, when the age was very clear on the page — it’s like they just randomly pick an index without even looking at the page or the data. I am not one of those. I will not submit a batch until I have looked at every entry.
That being said, most of the time, I don’t look very carefully at the data unless the indexers did not agree. If I notice one or both messing up certain things, then I will be more aware of those particular fields. Only once did I really look at every item on the page to verify the index because both indexers messed up so often.
I’m sure there are times when the indexers might curse me for changing what they typed. Especially when the two don’t agree, I do research. I will look up the names of counties and towns and spell them correctly when the enumerator clearly got it very wrong. When we don’t agree on a messily written name, I will check the 1930 census to see if I can find them there to choose which indexer is correct. Sometimes I ding both indexers because I think each is half correct.
One time, I only had to look at about three fields on the whole page. I wanted to tell the indexers “good job”, but there’s no way to do that. They’ll have to settle for their 99% on that one.
I have dinged indexers for putting the line number on blank lines, but it’s entirely possible that some of the indexers did that because a previous arbitrator “taught” them to do it that way.
Another arbitration involved the enumerator swapping two columns. What to do? One indexer typed what was written, which didn’t make sense, and the other “fixed” the columns. I felt bad dinging either of them, but one had to go and the decision (from someone higher up) was to type what you see. So eight people were born in “Same Place” and the city of residence in 1935 was “New Hampshire”.
Another arbitration had both indexers write “Canada French”, when the project rules say to just write Canada.
As an indexer, when I was arbitrated incorrectly (or thought I was), I clicked on feedback and asked it to be rechecked. From what I’ve heard, not much comes of that, but they might be watching a bit more to be sure some arbitrators are not messing up horribly.
The silver lining is that FamilySearch is working on an update so the database uses both indexer’s indexes. As arbitrator, I have sometimes marked both wrong, so I don’t know if they will completely eliminate arbitration, but at least both versions will be searchable. This will be most useful for those judgement call arbitrations, like on name spellings.
In the end, you just have to keep indexing and hope that you get a “good” arbitrator. And if you don’t, walk it off, have some chocolate, and get back in.
Or switch to arbitrating. But please be sure to read every page of indexing rules and be involved online somehow with other indexers and arbitrators so you have a better chance of getting things right.
After arbitrating a few pages, I went back to indexing, especially after the big pop-up notice on the project rules. But my arbitrator still got some of those easy rules wrong. So I’m back to only arbitrating. I had more fun indexing, but I have less frustration from arbitrating.
While I noticed one person stopped arbitrating because of all the complaints, it was my own complaints and possibly the others that make me want to arbitrate. Because we need arbitrators who have read the project rules. Because I’m tired of arbitrators who know fewer rules than I do, messing up the index. Because I signed up to contribute, and if indexing is going to make me mad, then arbitrating is what’s left.
And because I have more people to find in the 1940 US Census that I can’t find addresses for.
How much have you indexed or arbitrated this month?
The URL of this article is http://idogenealogy.com/blog/2012/04/29/indexing-and-arbitration/.