What I Like About Ancestry.com

[this post addresses Ancestry.com specifically, but the general concept applies to all the "big sites"]
 
I made one of my best genealogical discoveries on  Ancestry.com several years ago.

It was not in one of the "trees." 

It was not in one of the "hints."

That's not how I use Ancestry.com.

Since 1983, I had searched for "my" Ira (or William Ira) Sargent. Born in the early 1840s in New York State or Canada, he eluded me and appeared to have been dropped off by a UFO in Illinois in 1880, just in time for the census.  I had reason to believe that he spent time in Iowa or Missouri before 1880 for reasons I won't bore readers with here. I had found three or four other Ira Sargents who were contemporaries with my Ira in the Illinois-Iowa-Missouri area but had eliminated them as being "mine" for one reason or another. 

I pay regular attention to new sets of images of original records and finding aids on Ancestry.com that might help me find Ira or any of my other "problem" ancestors. I was delighted several years ago to see that a database of Iowa state census records had been released on Ancestry.com. I immediately searched it for Ira Sargent. There were several hits. All but one of the results I could "fit" into one of the excluded Iras.

All but one.

There was an Ira Sargent living in Davis County, Iowa, as a child in 1856. His details were consistent with "my" Ira and he could not be one of the "others." 

It didn't mean I had the right one.

It didn't mean I had evidence and it did not mean that I had a conclusion.

It meant I had a clue which I should follow. And I did.

The location of that census entry was the first crumb of information that lead to more crumbs of information which lead to more crumbs of information which eventually allowed me to connect "my" Ira to his family of origin. Most of those crumbs were not online and making the connections required analysis and thought. If finding Ira had been easy, I would have done it years ago. When I first found this Ira there were no "hints" linking his record to "my" Ira--that linking on Ancestry.com didn't happen until others learned of my discovery and put that information in their online trees.  Ira is one of those large number of ancestors for whom Ancestry.com hints don't work when no one else has found them. When I discovered him no one else had made the connection. (This also makes the point that if you don't want others to share "your" information, you shouldn't share with them in the first place.)

Compounding the problem is the fact that in 1856 Ira is not living with his biological parents. But that's another story and one I've written about in Casefile Clues. 

The indexes at Ancestry.com, while they can contain errors, do have the potential to bring new information to light. While the hints, the trees, and the indexes have their issues, there is no doubt that discoveries can be made. Just because discoveries can be made does not mean that we should be satisfied with the occasional database that is incomplete or the search that doesn't work quite correctly.

But making those connections is not an automated process.

One still needs to think. 

Cut off Cards

When a person doesn't know what a record is supposed to look like, it can be difficult to determine just what is missing. In fact, sometimes it can be difficult to determine if anything is missing at all. The World War I draft card below is for my uncle, John Goldenstein. It is missing something.

World War I Draft Card of John Goldenstein, Adams County, Illinois--obtained on Ancestry.com (http://www.ancestry.com), 4 December 2012

 When this card is compared to a "complete" card (see below), one can see the signature after item 12 on the left hand side. It appears that the right hand side of the card in the image above also has some issues. I've left feedback at Ancestry.com about the images (the problem appears to be with cards from Adams County, Illinois), but it is difficult to tell how long it will be before the images are fixed. The digitization of the cards from the microfilm is an automated process.

World War I Draft Card of Samuel Haase, Hancock County, Illinois--obtained on Ancestry.com (http://www.ancestry.com), 4 December 2012

The card above appears to be a complete card.

And it all reiterates the point of citation: not all versions of a record are created equally.

Missing Stones and Missing Pages

A recent blog post ("I've Never Been There But I Have Seen the Stone") discussed sketches of tombstones that appear in a Sargent genealogy from 1895. This book is available digitally in at least two locations, the Family History Library's website and Archive.org. Any citation to the digital form of the book should indicate  the actual site from which it was obtained. 

The pictures of the tombstones that were used to illustrate the blog post appear on a page originally inserted between pages 32 and 33. This page (and the blank one preceding it) are shown in these two illustrations from Archive.org's "page turner" viewing option. 

The first screen shot shows page 32 and the "blank" page on the right. The blank page on the right is actually the  back side of the page that has the pictures. Typical for the era, pictures were printed on pages that often did not have text on the reverse.

Hugh Sargent, of Courteenhall, Northamptonshire and his descendants in England, Aaron Sargent, Somerville, Mass. (1895)--obtained digitally on Archive.org (http://www.archive.org) on 3 December 2012, p. 32.

This set of images shows page 33 on the right and the un-numbered page that contains the tombstone images.

Hugh Sargent, of Courteenhall, Northamptonshire and his descendants in England, Aaron Sargent, Somerville, Mass. (1895)--obtained digitally on Archive.org (http://www.archive.org) on 3 December 2012, p. 32.

The image below comes from the PDF version of the book that was made from the item in the collection of the Family History Library. The view was made in Adobe Acrobat. There is nothing between pages 32 and 33. The picture either was removed from the book by a patron before digitization or it was missed. I am suspecting the latter. 

Hugh Sargent, of Courteenhall, Northamptonshire and his descendants in England, Aaron Sargent, Somerville, Mass. (1895)--obtained digitally in PDF format on FamilySearch (http://www.familysearch.org) on 3 December 2012, p. 32.

It never hurts to look at more than one version of any item, be it in print or digital format. There's a reason why we suggest that source citation include the precise venue in which the item was published, including the website and the date.

Not all digital versions may be created equally. Cite your sources.

This makes me think that there might even be justification for including which facility was used to reference a printed book as well.

How Do I Know I Have the Right Family?

How Do I Know I Have the Right Family? --this originally appeared in the Ancestry World Journal on 15 Dec 2006

by Michael John Neill

When families migrate from one area to another, it can be hard to
determine if you have really located the same group of individuals.
The difficulty is compounded if the last name and first names are
relatively common. It's important to be certain that the "true"
family has been located and that one has not mixed up families with
similar names. Male cousins bearing the same first and last name are
particularly easy to confuse.

One quick way to track families in the post-1850 era is through
population census schedules. The listing of all household members
facilitates the matching process, and every-name indexes make the use
of these records much easier than before, especially when the
residence is not known. However, the first close match on the list of
results is not necessarily the correct family. All matches to the
search terms should be analyzed and eliminated based upon what is
known about the family. What appears to be the "right" entry must be
compared in light of other records to determine if there really is
consistency.

On the other hand, searches of databases must not be overly strict,
as this can sometimes eliminate potential matches. I generally
perform a variety of searches for individuals I am seeking, including
some using Soundex and wildcard functionality. Sometimes it is easy
to determine if the correct person has been found (the person has an
unusual name, the person is living in the right location, names,
ages, and birthplaces of family members match up, etc.). Other times
it is not possible to make a definitive decision that the desired
person has been located. When families are eliminated, the researcher
should keep notes as to why these families were stricken from
consideration.

Generally speaking, when searching online census databases it is
helpful to track the type of search that is being performed as it is
being performed. Key elements in this tracking are:

• The first and last names that were put in the search box


• Whether a Soundex option was used


• Whether a search was performed with wildcards
• What year of birth was used (and what range of years)
• What birthplace was used

Reasons for tracking the search include:

• It is impossible to effectively modify an unsuccessful search when
  one is not certain how one searched originally or how one searched
  last week.


• Searching the same static database in the same way will typically
  produce the same results.


• It is impossible to remember each combination of search techniques
  that was applied. The "correct" combination will always be
  overlooked. It's Murphy's Law applied to genealogy.

One quick and easy way to track your online searches is to make a
spreadsheet with column headings for the various search boxes for the
database being searched. This spreadsheet can be printed and written
on while searching or those who are adept at toggling between
computer windows can fill out their chart as they search. Personally,
I prefer to fill out my chart of searches before I search, making
certain no combination of terms was eliminated. Then I can use the
chart to make certain I have conducted all the desired searches.

The Brices

An earlier column mentioned the family of William and Anne Brice and
how they were tracked in census records from Illinois to Kansas to
Missouri between 1860 and 1900.
As an example, let's
look at how their entries were obtained and what leads me to believe
I have the same family in four separate locations over four census
enumerations.

Generally speaking, census enumerations on any family should not be
viewed in isolation. (Space considerations do not allow us to include
the complete analysis in this column.) Rather, other records should
be utilized in order to determine if the tentative family structure
and migration paths are supported by other documents. Wherever
possible, obtain maps of all relevant areas to assist in viewing the
family's overall migration path. Search for reasonable alternate
spellings before assuming the "actual" family has been located and
consider if there are alternate situations that could explain the
records that have been found. We should search to see what is found,
not search to prove an already determined conclusion.

1860 Census-Ursa Township, Adams County, Illinois

William Brice, age 21, born Ireland, married within census year

Anne J., age 22, born Ireland, married within census year

The reference to the marriage within the year caused me to search the
Illinois State Marriage Index. An
index entry appears for William Brice and Ann Jane Belford indicating
an April of 1860 marriage. It seems very reasonable that this is the
same couple, especially since there were no other marriages in the
index for a William and Anne Brice (or any reasonable spelling
variant).

1870 Census-Chili Township, Hancock County, Illinois

William Brice, age 34, born Ireland

Ann, age 33, born Ireland

William, Jr., age 6, born Illinois

Mary A., age 4, born Illinois

Robert, age 1, born Illinois

Chili Township in Hancock County, Illinois, is close to Ursa Township
in Adams County. The ages of William and Ann are consistent with the
earlier enumeration. In both cases, William is a farmer (it is
important to note any extreme inconsistencies with occupation as
well). The ages of the Brice children are consistent with an 1860
marriage. The initial census search was conducted for a William Brice
(and Soundex variants) born in 1838 in Ireland, plus or minus five
years.

1880 Census-Bruno, Butler County, Kansas

William Brice, age 45, born Ireland

Anne J., age 48, born Ireland

William, age 16, born Illinois

Mary, age 14, born Illinois

Robert, age 11, born Illinois

Sarah J., age 9, born Illinois

James, age 6, born Illinois

John, age 2, born Kansas

The family structure is consistent with the 1870 enumeration. The
ages of the parents are off slightly from earlier enumerations, but
not so far off as to warrant any special concern. The initial census
search was conducted for a William Brice (and Soundex variants) born
in 1838 in Ireland, plus or minus five years.

1900 Census-Grant Township, Caldwell County, Missouri

William Brice, age 62 (born March 1838), Ireland

Ann, age 62 (born March 1838), Ireland

Jno. H. M., age 20 (born Mar 1880), Kansas

An unexpected move of the family. However, this was the only "match"
using our previous search terms that came even close to our desired
family. Anna Brice's death certificate (obtained via the Missouri
State Archives website) indicates that she was born in Ireland on 28
March 1836, the daughter of Daniel and Mary Jackson Belford. This is
the same maiden name for the "known" Anne Brice, wife of William.
Further research needs to be done, but it appears I have the same
family.

Wrapping It Up

• Perform searches that are not overly narrow so that close matches
  (which maybe the right family) are not overlooked.


• Constantly review information in light of already known
  information to be reasonably certain the same family has been
  located.


• Track what you do, so search terms can be modified as necessary.