Saturday, August 05, 2006

Advice for transcribers

Hello folks

Recently I asked two of our most experienced checkers for some general advice for transcribers. The second checker's comments are in brackets. They have both just finished very large 1851 pieces.

1. Since the programme (Excel) has a memory function, they should be careful using it. The transcriber must have spelled 'Louisa' as 'Lousia' the first time and never bothered to look at what he/she was typing after that. In the same way 'Ellen' showed up as 'Ellen Ann' after that.

2. Do use the resources of the net, et cetera when coming upon a strange name. I constantly had to go to 1841 or 1861 or 1881 to make corrections of names that were not all that difficult to read. [Michael: I use Google a lot, it will often give hits on place & personal names]

3. Be aware of gender - I often had a Francis, son, given as Female. (I find that if the transcriber misreads a name, then the rest of the tx'd data is often wrongly adjusted to match the misreading eg relationship & gender.)

4. Check birth towns. There is no Bugle in Cornwall but there is a Breage. [Michael: Six miles from where I am sitting is a village called Bugle!!!!!] (I just keep the GENUKI big/eng page open and refer to that. If it says Bugle CON, then it'll get checked as Bugle CON.)

5. If a transcriber comes upon an unreadable name and three lines later the same name is now written more clearly, go back and make the correction to the name instead of leaving the job to the checker.

6. If a transcriber wants to type in a wife's occupation as 'miner' or 'fisherman', he/she should realize that the occupation has probably been transcribed to the wrong person. (yes ... but if it's what the enumerator wrote (as is sometimes the case) then so be it.)

7. Do use software, where possible, to improve the quality of the original page - I use Adobe Photoshop to sharpen and improve the contrast. (Never yet needed to do it, but agree. With a flat screen you can often view it off angle to achieve some change of colour/contrast etc.)

(8. Quite often I find lines are transposed or translated mid-line (maths term ie shifted) ... I'm guessing that the transcriber is going down the page in columns, and I'm checking left to right. But it is a real pain with WINCC to put the lines back together again.) [Michael: This might be my fault - as I often advise transcribers to transcribe vertically]

(9. Also it might seem obvious but married couples usually tend to have similar ages with children aged at least 20 years less than the younger of the two. Yes I know there are exceptions, but it does beg a double check when transcribing.)

(10. Servants, lodgers, boarders, visitors, in-laws etc usually have a different surname to the main household.)

(But I do pity anyone battling with the fiche scans of 1851. They're very variable. Ancestry's done a much better job scanning at least, though their index is a little fanciful at times.)


JuliaM said...

Hi Michael - Now I'm concerned. Did the checker change all the Bugle references to Breage? If so, that's a ton of changes to be made. And what about the REAL Breage birthplaces in the same pieces? Zounds, hope you straigtened it out. (I have a vested interest, you know.) However, for everyone out there, I check maps, especially, for place names. Works well if the places still exist.... Julia

Chris said...

Hi Michael & All - the Gazetteer unility on GENUKI is very useful for pinning down place names because you can search by "beginning or end of word". See