One of the most important messages on this site is that you must always be aware of the limitations of the records you use, and the possibility of errors. One of the first pages posted on this web site in 2001, the Dangers of Internet Genealogy, pointed out that one of the biggest problems with online genealogy is it that it makes it very easy for lazy family historians, who never check anything to bulk copy erroneous research. The important thing is to realize that everyone makes mistakes and misunderstandings. No-one is immune. In 1635 a clerk (undoubtedly rather bored with the job) recorded the passengers sailing to America on the "Hopewell". Some came from the Hertfordshire village of Stanstead Abbots - but he didn't hear what was said and wrote down Stanstede Abbey. As a result of a query some 5 years ago I investigated The Myth of Stanstead Abbey - and how hundreds (and possibly by now thousands) of Americans had zombie-like copied other peoples family trees without anyone stopping to think whether such an Abbey existed. Some people went as far as to say their ancestors had been baptized by the Bishop of Stanstead Abbey, not realizing that if the abbey had existed it would be headed by an Abbot and not at Bishop.
However the commonest errors arise in documents such as the census - where many errors were made when the data was recorded, with further errors made at the time when the household forms were copied into the enumerators register in the censuses before 1911. Further errors have been made in the transcription and indexing - and almost every week I report such transcription errors. Over enthusiastic searchers can also make errors, especially when they over-rely on the indexes and don't check the original documents. There can also be difficult situations which are well described in Right Name, Wrong Body..
And of course there can be mistakes on this web site - and as soon as they are spotted they are quickly corrected. Some are simply typos - probably due to tiredness caused by spending too much time at the keyboard - but whatever the reason, it is important that if you discover a mistake you take action to correct it.
Richard has kindly brought to my attention a problem with my reply to Chris in 2010 about HARDING-YOUNG of Barnet in which I noted the marriage of John H Young to Bertha Folkard and then went on to suggest an identity for Bertha Folkard and at the same time indicated that Chris needed to buy two birth certificates and one marriage certificate which (as I now know) would have shown him that I had the wrong Bertha Folkard. I also suggested further checking of census returns which could have clarified the situation. The original text has now been replaced by that provided by Richard.
This is a good example of the Right Name, Wrong Body trap. In all the census returns between 1841 and 1911 there are only two Bertha Folkhard's listed - and they were born within 3 years of each other. FreeBMD shows a third, all three being born within 8 years, and this was followed up with 2 matching marriages and one death. One gets similar results from Ancestry and familysearch. I can't be sure what went wrong, but bearing in mind that some of my online answers go back to 1998 there is a trap that could catch anyone. Every year more and more information comes on line - and as "new" information comes on line much of it will confirm what we already "knew" but some may suggest other interpretations of the evidence.
The raises a question that has been worrying me for some time. About three quarters of all the answers on my site are more than 5 years old and some look very dated because they could only make the best use of the information that was accessible at the time. I sometimes look at some of the older answers, many of which refer to getting information from books or CD when better information can be got in minutes online. Perhaps it would be over the top to automatically delete all answers over 5 years old - but perhaps I should put a warning message on each answer to make it clear that much more information may be available online since the answer was drafted.