It's a challenge we often face in genealogy. You receive a microfilm full of records arranged alphabetically and have to figure out how far to wind through the reel to get to the desired record. I've never heard anyone give guidance on how best to search.
I think most of us look at where the film begins and ends and then wind through the reel until you're at what you judge is about the right point. You check and wind a bit, or a lot, forward or backward, then check again. You keep on doing that, making the moves progressively smaller, until you find your record.
What do you do when the file is digital, with a sequence number and name listed only for the first and last record? That was what I was faced with when I signed up as a beta tester for the Pilot for FamilySearch - Record Search mentioned in a recent post by Randy Seaver. FamilySearch issued two challenges, the first was to figure out how to sign up, the second a search challenge -- finding a specific person's record in the US WWII Draft Registration Cards. They are arranged alphabetically by personal name but only roughly indexed, by blocks of about 5,000 records.
There must be an optimum way to search, involving some arithmetic, based on the fraction of the way between the extremes of the names in the sequence your name is likely to be found. If the sequence has all the surnames beginning with O, and the name you seek starts with Ow you might start 23/26ths of the way through the sequence numbers . The name you find could could be "Overy, Charles", so you refine the arithmetic using that name and sequence number.
It would be really nice if a tool could be made available to do the calculations for you, maybe even one that accounts for the relative frequency of names in the region covered.
There's another approach to save doing advanced arithmetic -- the half-way technique. First check the record half way through the sequence, you still need to do a little arithmetic but dividing by two is something most of us mastered a while ago! If the record you seek is beyond that point you go half way again to the end (75%) and check again. If that's past your record go back half way (66-2/3%) to the middle. You can estimate the number rather than calculate. It helps to record your results so you don't forget what was where. When you're in the right vicinity just search sequentially forward or back to your record.
I found the record I sought quite easily, without the frustration of a cranky microfilm reader. The cards were easy to read, the viewer worked quickly and had a good zoom capability. I'm looking forward to the increasing availability of microfilms from the LDS collection online. Try it yourself -- the instructions to register are at Randy's posting.
Did you get a chance to see the screen shots accompanying my evaluation of FamilySearchLab's record viewer?
ReplyDeleteI am concerned about source documentation.
Myrt