Friday, July 20, 2012

1940 U.S. Census Index Comparisons - Post 6: McKnew in California


I proposed a methodology for evaluating the accuracy and completeness of the 1940 United States Census indexes on Ancestry.com and FamilySearch.org in 1940 U.S. Census Index Comparisons - Post 1: Methodology. 

In 1940 U.S. Census Index Comparisons - Post 2: Carringer in California, I displayed my comparison tables and found that Ancestry had 8 more Carringer entries than FamilySearch, and my judgment was that Ancestry was more correct on 7 of them, and FamilySearch for one.. 


In 1940 U.S. Census Index Comparisons - Post 3: Seaver in California, Part 1, Part 2 and Part 3 I displayed my comparison tables for Seaver entries in California, and found that FamilySearch had 22 more entries (147 entries) than Ancestry.com (125 entries).  I analyzed all of the Ancestry unique entries, plus the ones that disagreed but were indexed by name correctly.  My judgment for accuracy tally in these posts was: Ancestry = 13, FamilySearch = 29, Both wrong = 4, No decision = 7.


I want to move on to another surname in California - McKnew.  This is my wife's great-grandparents surname, and her grandmother's maiden name, and I think this study may be interesting because of the "Mc" prefix to this surname.

Here is the top of Table 1 for this group: 


There were 12 entries for an exact search for "McKnew" in California on Ancestry.com, and all of them were indexed the same way by FamilySearch.  However, there were 23 total entries on FamilySearch.  Here is Table 2that lists the six family units that were indexed as "McKnew" on FamilySearch but not on Ancestry:



The discussion on each of these is:

1)  George McKnew (ED 15-43C, page 64A) was indexed on Ancestry as "Mc Knew."  


The complete surname is McKnew...it appears that Ancestry.com indexed this with a space between Mc and Knew.  FamilySearch has the better index entry IMHO.

2)  Alice M. McKnew (ED 60-1050, page 2A) was indexed on Ancestry as "Mc Knen."


The "w" looks like a "w" to me - there is an 'n' three letters over.  FamilySearch has the better index entry.

3)  Alfred H., Gertrude, C. Rodney, Mari Jane and Richard A.McKnew (ED 38-498, page 6A) were indexed as "Mc Knew" on Ancestry.


Again, there is a space between the c and K. FamilySearch has the better index entry, IMHO.

4)  Alfred and Alice McKnew (ED 38-557A, page 61B) were indexed as"Mc Knew" on Ancestry.


Similar to the last one, FamilySearch has the better index entry.

5)  Anna McKnew (ED 38-579, page 62A) was indexed as "Mc Knew" on Ancestry.


Another one where Mc and Knew were separated. FamilySearch has the better index entry.

6)  Harry G. McKnew (ED 42-1, page 9B) was indexed as"Mc Iven" on Ancestry.


Besides putting the space between Mc and Knew, the Ancestry indexer read "Knew" as "Iven."  There is an I in the first name of the person above that should have been a clue,and the 'n' at the end was in the surname above.  FamilySearch has the better index entry.

7) In addition, the birthplace entry for Frank McKnew (ED 60-1070, page 9A, #1 on the Ancestry list) was indexed as "Washington" while FamilySearch indexed it as "DC."


Clearly, the FamilySearch index entry is more accurate.

To summarize, there are 12 differences found between the Ancestry.com index entries and the FamilySearch.org index entries for the McKnew surname in California.

My scorecard for the differences and the one indexing error is Ancestry = 0, FamilySearch = 12.  

Ancestry apparently indexed 12 entries for this name in California with "McKnew" and another 9 as "Mc Knew." They also indexed two entries that spelled "Knew" wrong.  My guess is that Ancestry.com did not have rigorous standards about names with a prefix like "McKnew."  I think that FamilySearch instructed indexers and arbitrators to not put a space between the prefix and the name, and a search for "mcknew"and"mc knew" provide the same matches (also for McNally, Fitz Randolph, O'Connell, etc.).

I think that I've done enough of these comparisons so that I can draw some conclusions, which I will do early next week.  

The URL for this post is:http://www.geneamusings.com/2012/07/1940-us-census-index-comparisons-post-5.html

Copyright (c) 2012, Randall J. Seaver

3 comments:

Anonymous said...

Just curious, Randy. Did you include other states besides California? ~Sharon

Patti E said...

Randy, as an indexer (for Family Search Indexing, or FSI) who has debated the space vs no space in Mc names issue, I found this very interesting. The official "rule" is what we call TWYS- "type what you see." My problem with this is, it is often a matter of interpretation whether the name has a space or not. And, as you have pointed out and I had thought, McPeople like their names with no space. I have "gone on record" as they say, as having decided that all McNames shall be indexed with no McSpaces. I'm considered a bit of a rebel, or loose cannon by some of my fellow indexers, but I can live with that.

The reason the powers that be are not worried about it, though is that the FS search function returns results both with and without the spaces, whichever option you put into the search box. That is to say, a search for Mc Name and McName will return the same lists in Family Search.

Just for your information, in case you are as curious about things as I am, the FSI protocol has two indexers transcribe each record, independent of each other. Those both go, together, to an arbitrator, who has to be the tie-breaker on any entries where the two indexers disagree. A good arbitrator (or indexer, for that matter) will not just try to decipher the writing, but will try, through searching on FS or Ancestry, to find the people in question in another record, where their name is written more clearly, in order to determine what their name actually is. I began indexing after finding records where my family members' names were so mangled as to defy credibility. One day I thought to myself, "I can do better than this!" at about the same time there was one of those little ads about indexing the 1940 census right before my eyes. So, I did. We have groups on Facebook and Skype where one can have other people look at difficult records, to have another interpretation of what is written. We think its fun!

Unknown said...

I concur with Patti E. I have indexed/arbitrated over 15,000 1940 Census Records. When arbtirating, I was disappointed when I had to choose between two correct name spellings - one with a space and one with no space. I usually chose nospace McName(s), but chose spaced De Name(s). FSI guidance was TWYS - but add the space if it was obious. I think the ultimate key point is can one find the record when searching - space or no space.