When we analyze a record we need to really examine it. Closely. Very closely. Here is a good example of why we should do that.
I've been wandering around in Ancestry trees today looking at trees that have my 3rd great-grandparents Stephen and Rachel Hopkins of Hancock County, Tennessee and later Harlan County, Kentucky. I keep seeing the same mistake over and over again. Almost every one of the 40+ trees I checked have given them a daughter named Ellen who was born about 1838. The problem is they did not have a daughter named Ellen. Unlike many of the other errors floating around the web, I can actually see how this one got started.
Let's take a look at the 1850 census for Stephen and Rachel's family.
See that 12 year old named Ellen? It certainly looks like Ellen at first glance and if you had no idea about Stephen and Rachel's children you might not have any reason to doubt that this was Ellen. When I first saw this record, I already knew that they had a daughter named Eliza (my 2nd great-grandmother) and I was confused as to why she wasn't listed with them in this census - until I looked closer.
Start with the 2nd l in Ellen. That's actually the tail of the y in Sally listed on the previous line. If you look even closer, you will see there's a dot in the middle of that y which indicates the 3rd letter of the name that looks like Ellen is really an i and not an l. The next letter that at first appears to be an e is actually z - see the tail of the z between the o and h in John on the line below? And that last letter that could have been n is obviously an a when compared to the n in John. The name is Eliza, not Ellen. But don't stop there. This could still be considered questionable so let's move on to the 1870 census.
Of course, nothing is ever easy so this record is a bit hard to read but the 4th person in this household is 23 year old Elisa. Apparently the census taker didn't like the letter z because her sister Elizabeth listed above her is spelled Elisabeth. And there is yet another point of confusion with this family. Some people seem to believe they couldn't have had daughters named both Elizabeth and Eliza (but they did, there they are in the same census record). Many of those trees with Ellen in them have Eliza's children tied to Elizabeth (who by the way went by Betsey and had two children of her own who are rarely seen in Ancestry trees but I've personally met some of their descendants).
Everything is not always clear at first glance so we really need to pay close attention when analyzing records.
More of Eliza's story can be found here.