Add a Transcription to Improve NARA's 1950 Census Index

Waiting for the big genealogy sites to index the entire 1950 US Census, I've been using the US National Archives site to locate my ancestors in that Census. Thankfully, NARA released a preliminary name index along with the images of the Census sheets, allowing for both name search and browsing. 

I've been most successful in locating ancestors when I try my search by inputting the state, county, Enumeration District, and surname of head of household.

Searching for Carrie

Above, results of my search for hubby's great aunt in ED 88-26 of Wyandot County, Ohio. Her name was Carrie E. Traxler. My search was for Traxler Carrie because the automated systems transcribed what they "read" on each line, and head of household was listed Surname Given. 

NARA's system can match on creative spelling, and the first result was the ancestor I wanted--even though the transcription wasn't perfect. Her name was NOT Traller Carriee, as the system read the enumerator's handwriting. But it's close!

Adding a transcription

NARA welcomes the public's assistance in adding correct transcriptions. It will accept both transcriptions that correct what the automated system says AND transcriptions that correct what the enumerator wrote. Note that this will not alter the actual Census in any way. It will improve the search system so others can find ancestors on the NARA 1950 Census site.

After I made a note of the line number where my husband's great aunt appears, here's how I added a transcription to correct her name as transcribed by the automated system.

First, I clicked the button to "Help Us Transcribe Names" and entered my email. NARA sent me a six-digit verification code, which I entered so I could begin my transcription. 

Next, I used the NARA drop-down menu to indicate line number...which took me to a blank of the screen shown above. I typed "Traxler" for last name, "Carrie" for first name, and "E" for middle name--all of which are shown in cursive handwriting on the Census form, but not accurately read by the automated system.

Other family members with same surname in same HH?

After submitting a transcription, NARA said thank you and invited me to "Add a transcription to the next line."

If Carrie Traxler's household included a husband or child or anyone with same surname directly below her name, I would click to add their surname.

This is a key step for families that share a surname and were enumerated in the same household. I didn't do it for Carrie, but I did do it for many other ancestors who had folks with same surname in that HH. 

Otherwise, the search system can't find these people by their full names. Remember, they would have been enumerated with a dash for surname, followed by a given name. My transcription replaces the dash with the actual surname.

"How do you spell that?" is the #52Ancestors prompt by Amy Johnson Crow for week #15.

Turn on Your GPS and Look at the Actual Image

Oh, it's tempting to accept the transcription here, and not look further.

After all, Aaron Work (1837-1924) is only a roomer in somebody else's household. He's a 1c4r, not a major figure in my husband's family tree.

But maybe he's rooming with a member of the FAN club (friends, associates, neighbors) possibly meaningful to family history?

Time to turn on the GPS (Genealogical Proof Standard). Don't settle for somebody else's transcription.

Always look at the image of the actual record. 

Here's what happened when I went looking for Aaron Work with the GPS in mind.

Transcription says household of King family

I was researching Aaron Work for last week's #52Ancestors prompt, because he was a fire insurance agent and the prompt was fire.

In the 1920 US Census, Aaron was widowed and a roomer, as the transcription shows. Following the Genealogical Proof Standard of referring to the original record rather than relying on a transcription, I clicked to view the Census page.

And that's when I saw something that I've heard about but not yet experienced in 23 years of genealogical research.

"Supplemental" entry

Aaron's entry was added later by a supervisor, long after the enumerator had completed that page. It was marked as "supplemental" and added on April 1st, whereas the rest of that page was dated January 22.

The official Census Day in 1920 was January 1st, but officials continued to follow up and look for people who had not been counted in the first round. Apparently Aaron was one of those people missing when the enumerator came to his door.

Look at the red arrow on the supplemental entry above. The note shows where Aaron really belongs in the Census: "See 4A, line 37." Also note that Aaron has many blank lines above his name. The entry directly above him, with blank lines in between, is of the King household, and therefore the transcription seems to have lumped him into that household. Wrongly, as it turns out.

Page 4A, line 37 

I clicked backward on the Census images from page 8A, where the supplemental entry was listed, to page 4A. Above is a snippet showing number 37 at far left of the Census page. The date of this page was January 13, 1920.

The two people in this household are brother and sister, names that I don't recognize but will have to research to determine if they're at all related to the Work family.

Where Aaron Work was on the day the Census enumerator originally came around, I can't guess. I only know that he was later tracked down and added as a supplemental entry.

If I had accepted the transcription without checking further, I could have been chasing King as a potential FAN club member--and gone down an entirely incorrect path.

Thanks to the GPS, I didn't take a wrong turn. I looked at the image and saw the household where Aaron Work would have been enumerated if he had been home.

In citing my source, I need to mention both the 8A Census page of the supplemental entry and the 4A Census page of the household where the supervisor said he resided in 1920.

Lessons Learned in My Virtual Research Trip

Today, when I was clicking my merry way through online records pertaining to my husband's Slatter family, I discovered one shortcut and was reminded, yet again, of the value of checking originals.

Above, the shortcut I found to cut through the clutter of hints. My husband's Slatter family tree on Ancestry has more than 9,000 outstanding hints. Most of those are for ancestors too distant to be a priority. So I clicked on "records" to choose only those hints, then brought up the "filter by name" sorting option. (The default is "most recent" which means when Ancestry added that hint.)

By entering "Slatter" in the surname search box, I was able to view only record hints containing that name. Of course, I could have searched by first and last names, but given the creative spelling in so many records, I wanted to click through all Slatter record hints individually. Focusing on one surname enabled me to make progress, rather than being sidetracked by hints unrelated to my current research.

Now for the reminder about original records vs. transcriptions. The three dates on this record of marriage banns from a London church are 1 Dec, 8 Dec, and 15 Dec. The handwriting is very clear. At top of the page, not shown here, is the handwritten year--1907. Yet the transcription of this record says the year is 1908.

By reading the handwritten record, I was able to enter the correct dates for the marriage banns of Thomas Albert Slatter and Jessie Alice Elms. Also, the marriage license original confirmed the actual wedding day as 28 December 1907.

It's never safe to assume a transcription is accurate, let alone complete. It took only a few more clicks to view the originals and extract every possible data point.

My starting point for today's post was Elizabeth O'Neal's Genealogy Blog Party, July edition: Virtual Research Trippin'.