I have been a researcher for almost 20 years. I have volunteered as a transcriber. It is not easy. Some of the documents were in bad shape when they were digitized. Some of the images are poorly scanned. Sometimes the writing is small, or a very fine point pen. Other times, the ink fades, paper yellows, and you can't make it out. And none of that even accounts for the fancy handwriting.
Transcribers do their best, but sometimes mistakes happen. And sometimes, the mistake was in the original. I'll you two examples.
My grandmother, Elsie Florence Stadtler, has two birth certificates on record. We discovered this after she passed away, so no one really knows why. The first one, filed a few days after her birth, shows her name as Elfie Florence Stadtler. The second, filed about five years later, is marked as a correction, lists her name as Elsie, and her birth year is one year older. So we asked ourselves why they filed a new birth certificate, and decided it must have been required for her to attend school. The only confusing thing is the change in her birth year, because we always celebrated the date from the first certificate. So now, we have two birth certificates with mistakes. By the way, these are for the same person because the parents match, and there are no other records in all of Chicago with the same names, and Else only had one brother.
The second example is a recent project I was working on for a friend. While searching for census records for her ancestor's family, I found two listings for him in 1910. One database entry shows him as head of house, married, several children, father-in-law, and aunt. Great! This is a jackpot!
Second listing didn't concern me too much, because it's a sort of common name, but what struck me was that the name I was looking for was listed as father-in-law, while the wife and children had the same name as my search, and the head of household had a completely different last name.
I checked all the fields, and sure enough, the roll number, page number, Enumeration District, were all different. I had no choice but to study the original images. (I always do anyway - there's some really juicy information in there!)
I was shocked to find that the transcriber had jogged a line in the record and pout the wrong last name for head of household. Not only that, the father-in-law's name was also written in the census correctly, and transcribed incorrectly. Additionally, the aunt was described as "aunt-in-law" in both cases, and all first names were correct, and in the same order.
While I'm sure it didn't happen often, I am 99% sure this family was counted twice, which I wouldn't have discovered if I hadn't checked the original images. By the way, reading the originals also explained, at least partly, why they were at a different residence the second time. Dad changed jobs - both of his jobs could have offered housing, so they probably moved due to the job change.
Always check the original images to verify the transcription.
