I've been doing new research, and redoing old research, for my book about John Owens, Indian Trader, and his family using online sources. When I began my research in the late 1990's I used mostly microfilm from the Family History Library ordered from my local Family History Center. Using the indexes on those films or ordering separate microfilmed indexes, and sometimes just paging through every page of a microfilm, I was able to find the information I needed for my research. Now my research is nearly exclusively online.
Here are some of the problems I've encountered in my research online recently:
Ancestry.com's index of a family death certificate contained errors. When I first saw this information at Ancestry I wasn't sure this was the correct person? When I saw the death certificate it was the correct person. Josephine Owens' father's name was indexed as Janus Owens, which actually should be James Owens. Her husband's name was Benjamin Durham, but it's indexed at Ancestry.com as Guy Darrow. You can't count on indexed names being correct.
The person who filled out the form had handwriting that was difficult to read. I however wouldn't index the name of her husband as Guy. It does look like Benj.. When I saw his name recorded as Guy Darrow I thought she remarried, but actually she was still a widow when she died and she didn't remarry.
The actual death certificate contains multiple mistakes. The informant was the son of Josephine. He seems to have given his grandfather James Owens name as his mother's father, instead of her father William F. Owens' name. The document also states Josephine died on 9 February 1926, but the doctor last saw her alive on 10 February 1926. Anything written down can be in error. That's why using multiple sources is important.
Another problem searching for census records involved the 1850 Census. Again the handwriting of the enumerator is difficult to read. I looked for Robert S. Owens online using the search fields for census records at all of the sites with census images. I could not find him, but I knew he was living in Bracken County, Kentucky in 1850. I finally stumbled upon the reason I could not find him. His name was indexed as Owand, instead of Owens. If you already know this man's name is Owens you can decipher this difficult to read handwriting. I'll be adding the name Owand as a variant spelling of Owens to my list.
Another problem I've encountered is poor quality unreadable online images. This is the 1860 Census for Effingham County, Illinois for my Owens family. As you can see it's tough to read the names, but if you already know the names you can make them out. I could not find this record by searching for it. I had to use my microfilm copy to find it.
This is the same 1850 Census image for the Owens family. I made this copy from a microfilm. The names are more legible.
Here is another record that is even worse.
It appears Ancestry.com solved the problem of the difficult to read pages by leaving them off their site all together? I can't find page 1125 for Effingham County, Illinois?
I love the online records but now know to be extra cautious if records fail to surface when searching on genealogy websites. I did make a major find for the Owens family tree by searching FamilySearch deeds online. I found a new member of the Owens family named Julia Ann Owens born about 1810 in Kentucky. I also found a deed that confirmed what we suspected about Josiah Owens being the son of James D. Owens and Francis Watkins. The deed states Josiah was the son of Francis; leaving no doubt about his parentage. I could never afford to order all the films for every deed book necessary for research so the free online records have been so helpful.
Sometimes the locked records at FamilySearch are a little puzzling as in why are some of these indexes searchable while others aren't? The Philadelphia, Pennsylvania A-H Grantors indexes are searchable, while the I- Z are only searchable at the Family History Library or a Family History Center?
Sometimes you find the documents are badly damaged. I'm happy, however, that even the damaged records are searchable because who knows, someone may be able to decipher something from them?
Every site with census images has some that are illegible. I wouldn't assume that someone isn't on the census unless you've done a page by page search of a complete set of records. It's also possible a page of an online record could have been missed when the records were filmed. The fact you don't find something in online records doesn't mean it doesn't exist. Looking for other copies of records at more than one website, or offline, could help you find what you're looking for.
I am using this blog to record my genealogy research progress. My research at this point is concentrated in Tennessee, Virginia, Kentucky and Pennsylvania records. Since many of my ancestors ended up in the Midwest and West my research sometimes takes me in that direction. I also research Irish records.