{"id":1891,"date":"2016-01-23T05:58:20","date_gmt":"2016-01-23T05:58:20","guid":{"rendered":"http:\/\/rootdig.genealogytipoftheday.com\/?p=1891"},"modified":"2016-01-24T13:50:36","modified_gmt":"2016-01-24T13:50:36","slug":"no-sources-for-imagined-places-i-cannot-see","status":"publish","type":"post","link":"https:\/\/rootdig.genealogytipoftheday.com\/?p=1891","title":{"rendered":"No Sources for Imagined Places I Cannot See"},"content":{"rendered":"

[Note:\u00a0<\/strong>All images were current as of 23\u00a0January 2016 at 11: 30 pm central time.]<\/em><\/p>\n

I’ve had a little time to work with the “U.S., Obituary Collection, 1930 – 2015<\/a>” at\u00a0Ancestry.com<\/em>. And I’m not impressed.<\/p>\n

Complaint number one: What paper was the obituary in?\"obituarycollection\"<\/strong><\/p>\n

The specific newspaper from which the obituary “index” information was taken is not included in what Ancestry.com<\/em> calls the “record.” In some cases, the name of town where the newspaper was published is indicated. In other cases it is not. An “index” should lead the researcher\u00a0to the actual entry from which the index information was extrapolated. This database does not do that.<\/p>\n

How difficult is it to harvest the name of the newspaper when harvests information from the obituary?<\/p>\n

 <\/p>\n

Complaint number two: Grabbing any state or county for a place name<\/strong><\/p>\n

\"obituarycollection-towngrabbing\"<\/a><\/p>\n

Poor geographic “plucking” is another criticism of this database. This is what happens when instead of just extracting data from a published item, the search algorithm apparently generates missing information. \u00a0The location of Quincy is likely mentioned in the obituary of my great-aunt, Mildred (Irvin) Trautvetter as illustrated in the second image. The database “index” entry concluded that it must be Quincy, Massachusetts. It’s not. It’s Quincy, Illinois. I’m reasonably certain that her obituaries did not indicate she was married\u00a0in Quincy, Massachusetts. Newspapers in Adams County, Illinois (of which Quincy is the county seat) typically know where Quincy is located and I’m pretty certain the Trautvetters didn’t travel to the East Coast to get married.<\/p>\n

The other entry for Aunt Mildred has a similar problem. Again the algorithm apparently extrapolated in addition to simply extracting. What was probably a place of birth of Adams County, Illinois, became Adams, LaSalle County, Illinois.<\/p>\n

\"obituarycollection-towngrabbing2\"<\/a><\/p>\n

Just tell me what the obituary contains. When pieces are missing I can try and figure them out using my brain and not a piece of software.<\/p>\n

Complaint number three:\u00a0there appears to be information in the database that’s not showing up on what\u00a0Ancestry.com\u00a0<\/em> calls the “record entry.”<\/strong><\/p>\n

The first illustration in this post is the complete record entry for my uncle, Roger Neill. It was obtained by searching for “ida trautvetter” in the keyword box as shown in the illustration that follows.<\/p>\n

\"findignuncleroger\"<\/a><\/p>\n

The word “ida” and the word “trautvetter” do not appear in the complete record entry for Roger Neill (as shown in the illustration)-yet it is a search result:<\/p>\n

\"idatrautvetterserch\"<\/a><\/p>\n

I don’t\u00a0know<\/em> why. I had the exact box checked, so the word “ida” and the word “trautvetter” should have appeared in the record entry. Both those words appear in the two entries for Mildred Irvin Trautvetter (although not in close proximity to each other). My uncle’s obituary is a different story. Ida Trautvetter was his mother’s maiden name and that name likely appeared in every obituary for him. Yet does not show in the record entry for him even though that record entry came up while searching for those words in the keyword box.<\/p>\n

A search for “cecil” and “neill” in the keyword box also brought up Uncle Roger’s obituary (there were too many results to insert the screen captures in this post). \u00a0That’s the name of his father which also probably appeared in every obituary for Uncle Roger. A search for Uncle Roger’s wife’s maiden name (which I won’t post here because she’s still living) also brought up Uncle Roger’s obituary. That name does not appear in the record entry for Uncle Roger either.<\/p>\n

None of those words appear in the displayed record entry for Uncle Roger–except for Neill. Yet the search indicated that they did. \u00a0It appears as it the “entry” includes other words\u00a0that simply aren’t being displayed. I know that those names appeared in my uncle’s obituary.<\/p>\n

My Mother has two obituary entries in the database. Neither of the database entries include her mother’s maiden name. A search for that maiden name as a keyword pulls up my mother’s entries even those\u00a0Ancestry.com<\/em> does not include that name in the record entry. I know that Mom’s mother’s maiden name appears in her obituary.<\/p>\n

I shouldn’t have to guess why a search result came up when I am performing an exact search.<\/p>\n

I realize databases are clues and I realize that people should use information responsibly. However the reality is much different. How long beore these extrapolated locations\u00a0from these obituary “index entries” will appear in actual\u00a0Ancestry.com\u00a0<\/em>trees. Those trees may or not may even link to the index entry from this database, leaving the viewer with no idea where the information came from. And if this database is tied to the entry as the source of the information, it\u00a0only gives a date of publication of the original newspaper obituary and (if you are lucky) a city.<\/p>\n

I’m guessing that part of the reason the newspaper isn’t named is that there’s information in the database that’s not in the newspaper (the extrapolated locations).<\/p>\n

Be that as it may.<\/p>\n

An index that can’t tell me where the information was originally obtained is not really an index. Indexes are tools to direct researchers to the record from which the index entry was made. Simply saying the original information came from a newspaper published somewhere in the United States on a certain date is not specific enough. If there was only one newspaper published in the United States then maybe that would work.<\/p>\n

The extrapolated location information needs to be removed. Period.<\/p>\n

If there are additional fields that aren’t displayed in the record, I’d like to see them. Otherwise I’d like to know how exact searches turn up matches whose record entries do not have those keywords in them.<\/p>\n

In summary, my frustrations with the “U.S., Obituary Collection, 1930 – 2015<\/a>”\u00a0are:<\/p>\n