APG-L Archives

Archiver > APG > 2008-07 > 1216953171


From: Neal Underwood <>
Subject: Re: [APG] 1920 census going online - free
Date: Thu, 24 Jul 2008 21:32:51 -0500
References: <mailman.66321.1216766528.9525.apg@rootsweb.com>
In-Reply-To: <39786.50164.qm@web62401.mail.re1.yahoo.com>


What, then, is the point of FamilySearch putting an "index" to the 1920
census on its website and not accompanying it with the underlying schedule
images when they already have incurred the expense of scanning the images?
Do they expect the casual researchers to drive to an FHC or subscribe to
Ancestry to obtain complete, accurate data? Since images to other censuses
are posted free on FS, their intentions are confusing.


At 06:43 PM 7/22/08 -0700, Ruy Cardoso wrote:
> > Date: Tue, 22 Jul 2008 14:27:11 -0500
> > From: Neal Underwood <>
> >
> > Hopefully, the new "indexes" will contain sufficient detail and be
> > displayed in such a manner that referral to the original
> > image is only required for questionable or specific data items. For
> > example, I generally only use the transcription for the 1880 census and
> > only occasionally need to refer to the schedule.
>
>Nice idea, Neal, but still dangerous. I can see how relying just on such
>an idealized index would be tempting for any number of reasons, but why
>take that particular shortcut when the corresponding images are so readily
>available through a number of sources? And how can one assess what is
>questionable without making the comparison of index to image? None of
>this diminishes the usefulness of something like the 1880 census
>transcription; it's the excess and largely unnecessary reliance on it as a
>sole source that I'm questioning.
>
>I spent several years compiling an "index" along the lines you're thinking
>of (for certain parish records). Yet even with my intimate knowledge of
>the families involved, name spellings, handwriting, and so on, I'm sure
>there are still misreadings aplenty (and they are now enshrined in the IGI
>as well). Despite -- or perhaps because of -- my thousands of hours of
>working with these records, I would unequivocally advise users of my
>compilation to look at the original records rather than just relying on my
>interpretations of them. Other eyes may see things differently than I did.
>
>Perhaps as a little test you could take, say, a couple of dozen random
>entries from the 1880 census transcription and compare them to the
>underlying images. I've not done this myself, so I'm not sure how
>accurate you will find the transcription to be. But then I'm not the one
>relying so heavily on that transcription, and I would think such a test
>would be useful in evaluating your approach.
>
>Ruy Cardoso
>
>
>

This thread: