Monday, April 18, 2011

Finding Freiburgers

While reviewing my Freiburger ancestors recently, I was forced to confront a research problem that stymied me in the past: I was unable to find 1880 US Federal Census entries for three Freiburger families who lived in Boston in 1880. I was very confident that the three families were in Boston because of other evidence including vital records and city directories, and I found census entries for all the people involved in 1870 and 1900, but I could not find them in 1880 and of course, the 1890 census entries are not available. Everytime I visited one of those records in my TMG project, the 30-year gap between census entries was staring me in the face.

I use online genealogy sources a lot because what little time I have for genealogy research is late at night when libraries and other archives are not open. I've been using Ancestry.com for at least 11 years and I am a very satisfied user. In 2000, the site was not as fast or reliable as it is now, the user interface has improved since then, and there's a lot more data available, especially digital images of original documents.

Ancestry.com is noticeably better than the two other sites I use most frequently:

This post is mostly about the techniques I used to find the Freiburger families on Ancestry.com, but along the way I'll make some comparisons between Ancestry.com, FamilySearch and American Ancestors. I have loose plans to write an article that compares those three sites, but who knows if I'll ever get to it!

Finding Freiburgers can be difficult because there are many spelling variations. Some people spell it with an "ie" rather than an "ei", still others spell it with a "y", the "burger" part is sometimes spelled "berger", and transcription errors add to the fun. To increase my chances of finding Freiburgers, I use wildcards when those are allowed. On Ancestry.com, I typically search for "fr??b?rger" and "fryb?rger", and I try other patterns if those fail. Those patterns, and others I tried, didn't help me locate the missing 1880 census entries.

Late in 2010, after multiple attempts to find the 1880 census entries using wildcard searches, I browsed the images for the enumeration districts where I thought the missing families would be. I had addresses from city directories and I was reasonably confident I would find the census entries. My confidence was misplaced, and I didn't find any of the families. Even though I have now located them via more imaginative index searches, I am not sure why manually searching the enumeration districts didn't work.

The key element of finding the three families was to abandon the notion that a surname is an essential component of every online search on a list of names. Intellectually, I know that a surname is not required, but as a practical matter I sometimes operate on autopilot and enter a surname term, and then try variations of the term if the initial search fails. When I remember to think outside the name boxes, I've located many hard to find records using other search keys.

To use the technique effectively, you must choose something relatively unique about the person or family that is present in the database index and will help to reduce the size of the result set returned by the query. Using a single non-surname term may produce a result set that is too large to be useful, but each additional term will reduce the number of results.

Here's how I used the technique to find the Freiburgers in the 1880 census.

I'll start at the moment of inspiration. I assumed that the enumerator misheard or misspelled the surname, or perhaps the Ancestry index used a poor transcription, or both. I decided that I might never guess the correct surname search term to use, and around the same time, I noticed that the 1870 Census entries for the Freiburgers all specified a birthplace of Baden. I already knew the family was from Baden, but what occurred to me at that moment was that Baden was not a large country and there probably weren't that many immigrants to Boston from Baden. By itself, a search of people enumerated in Boston who were born in Baden would yield thousands of entries, but combined with some other limiting term, the birthplace might produce a small enough result set to be useful. I decided to use birth year as the other limiting term.

Anna Marie (Kappler) Freiburger immigrated to Boston from Baden in 1867 and she was enumerated with her husband Johann and their children in the 1870 census. Johann died in 1878, but Anna Marie and her children were still living in Boston in 1880. From other evidence, I had an estimate of 1820 as her birth year, so I opened Ancestry's search page for the 1880 census and entered the following search terms:

  • Birth Year: 1820 +/- 2 (Exact only)
  • Birth Location: Baden (Restrict to exact)
  • Lived In: Suffolk County, Massachusetts, USA (Restrict to exact)

When I include a name in the search terms, I typically turn off the "Restrict to exact" properties of the other terms to guard against mistakes in the index or mistakes in my assumptions or prior research. When omitting the name, however, I turn on "Restrict to exact" because otherwise there will be too many entries in the result set.

I was glad to see that only 52 people matched the terms, and even happier to discover "Anna Friedburger" as the 15th entry. One click later, I was able to confirm that "Anna Friedburger" was really Anna (AKA Mary) Freiburger, the widow of Johann, and the adult children I expected to find living with her were all enumerated in the same household. Success at last!

I was a bit disappointed to find that I had only myself to blame for being unable to find her using surname search terms. The transcription was accurate, and while the enumerator spelled her surname with a "d", that was not uncommon. I knew the name was sometimes spelled that way, but I had neglected to account for that in the search terms I used in previous searches.

I did a search using the surname Friedburger in hopes that surname was also used for Anna Maria's adult children who were not living with her. No luck with that, but reusing the same technique I used to find Anna Maria worked two more times, once for each of her adult sons who were not living with her.

Anna Maria's son Jacob Freiburger was found via the same terms as above except for an estimated birth year of 1855. I was less certain of his birth year so I expanded the +/- factor to 5 years. He was the 49th entry of 50, enumerated as "John Flyburg". Again, the Ancestry index transcription was correct, but the enumerator got it wrong.

I changed the birth year to 1846 +/- 1 to find Anna Maria's son George Freiburger. He was enumerated as "George Fayburger". In this case, there was a transcription error. The text on the digital image is quite faint, but I believe the enumerator wrote "Fryburger".

The technique described above will work on other sites as well as Ancestry. It typically works best when you can limit the search to a specific database within a site. For example, I found a death register entry for Bridget (Houghton) Freiburger in the "Massachusetts Vital Records, 1841–1910" database on American Ancestors using the same technique. Bridget's surname was transcribed incorrectly as "Freihurger"; I found her via a search for first name Bridget, death events only, year 1880, in the city of Boston.

The technique is a little more difficult to use when the genealogy web site does not have database-specific search pages. On FamilySearch, for example, you can filter the result set by category (where categories equate to databases or collections on Ancestry.com), but I have not found database-specific search pages. The search terms are the same for all searches, and that restricts the options you have for reducing the size of the result set. When searching for a census entry, for example, you can't limit the search to a specific census year, and you can only specify a single location. So, for example, you can't specify both a birth place and a "lived in" location as you can when using Ancestry.com's search page for the 1880 US Federal Census.