(all the on-line material can be accessed via www.manxnotebook.com/parishes/andreas/index.htm and thus I wont give indiv page refs)

Some time ago there was a discussion of a possible Manx 'IGI' avoiding manner of the errors that have crept into the existing IGI.
I spent a few days examining the problem to gain an estimate of the problem and time required to correct them.
I choose to look at Andreas as my initial examination of Parish Registers as found using batch numbers in the IGI indicated that this was probably the worse ( Michael, Onchan & St Marks are also affected badly others less so) .
It should be pointed out that all Andreas entries pre 1717 are copies - three such transcriptions were looked at (a)1911 General Registry transcription - this is also available on Mormon microfilm and is what most people refer to as the Parish Reg - the IGI entries that claim to be from the Parish Registers (ie the Cxxxxxx & Mxxxxxx batch numbers are actually from this)
(b) an 1827 copy of the PR now in Manx Museum (PR2) and
(c) the 1717 PR (PR1) in a worn state but still readable also in the Manx Museum - all entries from c.1656 Marriages, 1666 baptism etc in this are actually copies from a register described at the time as old and decayed and now presumeably lost.

I started by comparing the 1911 with the IGI entries that claimed to be from it - the first point was that the 1911 transcription was in general very accurate (not unsurprising as both PR1 and PR2 are very clear (at least for Marriages, PR2 is less well written for Baptisms), I found one or two minor errors e.g. Jony transcribed as Jany). However the 1911 transcription omits all entries between the 17 August 1669 and 1671 ! I looked then at PR2 and PR1 to see if these were due to a skipped page but could see no obvious reason - this was the first and unexpected discovery. I compared PR2 and PR1 - in general for marriages PR2 was very good and could in a few cases resolve doubtful readings due to wear in PR1 - I have transcribed all marriages in PR2 upto 1757 + any associated comments in the register (though those after 1685 are only on my CD ROM)

Turning to baptisms the 1827 is less accurate but the 1911 looks very good though in a few cases there appears to have been misreading due to the worn page - the 1827 copy can be used to guide corrections but is defficient in (a) the transcriber did not realise that year started on March 25 thus confusing pre 1753 entries,(b) the transcription modernises many family names (eg doubling some final t's) and (c) a few entries clearrly readable in PR1 are missing - usually because it or the next entry are somewhat different from normal entries.
I have transcribed all baptims upto to 1732 + a few additional years (again most are on my CD ROM) - I also transcribed the Confirmations upto 1760 (also on CD)

Unlike many other registers there were very few additional comments in the register - I was hoping to find a seating plan for the church as this would probably provide a key to understanding the order of 1757 Visitation census - however the 1760 transcript of the 1704 composition book provides considerable info.

The IGI for Andreas is corrupt - many years have had their 1911 transcipts removed and replaced with a collection of readings some of dubious accuracy (polite form of 'totally wrong'!).
The baptisms have both had years removed and replaced by doubtful readings (eg claiming birth dates which correspond to the christening dates in the PR, other having confused two adjacent entries, or incorrect years, invented months etc) but many Sayle and Moore names were also replaced by readings from a certain batch - many of whose entries I could not find in the PR
I suspect that entries post c. 1780 (marked as 1911 transcription) may be accurate and reasonably complete though I did not check this.
The time taken was about 1 day for every 25years of the Andreas PR (most years had about 40 baptisms) but checking against the 1911 transcript and adding the comments omitted from this may be somewhat faster.

Thus providing a clean copy of the PR independent of the IGI is certainly possible but requires some considerable effort

On the matter of illegitimate births - the PR for the period had two forms - one giving the family name as that of the father (where known) and another giving child of X & Y - in all cases the IGI has indexed under the mother's name even in the cases where the PR placed them under the father's name.