Is there a census working group or other plans to incorporate census records?

Question

Like Find A Grave, Census records can be used to track duplicates. Also we can use position on census records to find relatives. 

I am wondering if there are any plans to incorporate census records into wikitree in such a way as to isolate a one record / one person match. 

I imagine we can create a category for each census page. 

Take my grandfather Lawrence Martinucci. 

He appears on the census at 

United States Census, 1920

District	ED 33
Sheet Number and Letter	2B
Household ID	38
Line Number	84
Affiliate Name	The U.S. National Archives and Records Administration (NARA)
Affiliate Publication Number	T625
Affiliate Film Number	94
GS Film Number	1820094
Digital Folder Number	004964276
Image Number	01029

 

Perhaps a category such as 

1920 Census, California, Del Norte, Crescent City, ED 33, Image 1029

 

This way, we could see who else thinks that this is their ancestor. I have found many errors that have people linked to the wrong census records. 

Also imagine the helpfulness of being able to look at a specific county in 1920 and being able to get a list of individuals who appear only on this one census. Like a where are they now or where did the go group. Orphans. 

 

Comments

This is great.  

Maybe one day We Relate and WIkitree will talk and this will all happen like magic.

---

I love this. I floated the idea of a source-based genealogy platform some years ago, but WikiTree is horse and buggy on a superhighway for the purpose.

I imagine that with things like the hundreds of millions of census entries and billions of data points, it is going to have to wait for bots to entirely automate the system.

It seems that what is being proposed above is to create a sort of semi-automated hybrid onto WikiTree. More power to you. I hope the wizards can build an easy and useful tool of it.

---

I am guessing that this means there is a unique identifier on each name in the census that we can use.

If we could find that number/designator, all would be solved.

Simply a Census template with a single number/identifier.

---

Each sheet has a printed line number for each name box, Lance.

Some lines are blank. Such as at the end of a household on the bottom of the sheet, or when streets change from one to the next, as separators.

---

Yes. That is where I was going with asking if there is a working group that is creating a standard reference.

No sense in entering information in an unformatted way.

 

Referring to Magnus wiki-bot the unique identifier might be available from the Census Department. If we can somehow derive that number we would be set.

---

However, there are flaws in the original data. For example, I have seen households listed twice, in two separate locations.

This probably occurs when people move during the months, and also children shifted around,  between relatives, depending on need and life events.

So the system has to allow for more than one census match per individual.

---

Ooohps its not my bot ;-)

I have nearly no understanding of US genealogy and census record. If you once tested Swedish genealogy and church books you never turn back ;-)

---

When I find somebody in a UK census on Ancestry it can often point me to the same person in other censuses.  This is good stuff.

I think future-genealogy is all about tagging up sources with unique person-IDs - because searching by name is rubbish really.

Then the system builds "profiles" by retrieving all the sources tagged with the ID.

 

The Wikipedia {{cite census}} template has two problems

1 it doesn't create a link

2 the cited entry isn't Wyatt Earp, it's one of his lady friends discussed in the article.

---

And don't forget that the people who transcribe those records and the census takers themselves often make terrible errors in first and last name spellings

---

1) it doesn't create a link

As templates are just text substitution then its easy to change the template when you have a place to link to

2) the cited entry isn't Wyatt Earp, it's one of his lady friends discussed in the article.

Then you say samesas=no see G2G

Answer 1

Hi Lance!

I think overall we should somehow incorporate census records into WikiTree, maybe with a template of some sort or some other way.

But I think the categories would be too tedious for leaders, project coordinators and just profile managers to add.

Just my thoughts.

Comments

The benefits are create to have sources as templates. One problem is that when doing GEDCOM import you normally have a structure that is converted to just text when you import....

Maybe the simplicity is the best part of WikiTree.... you win something and you lose some....  I guess doing more quality in an efficient way then you need structure...

---

I was thinking more about the category idea and I came up with some problems already. One is that checking would be complex in that peoples names were often changed, misspelled or otherwise in error. When looking then at the profiles in the category, they will not match the names on the census. There would be too many steps to relate them to one another to make this a true solution.

Answer 2

That is quite an idea as it would also help people see who else from that census page is on WikiTree..

I do agree with Sarah in that a template might be the better option but would need to be thought out more to figure out what would be the better option for this. :)

Thanks for bringing up this idea!

Answer 3

Hi Lance,
Yes,  we have a census project started: https://www.wikitree.com/wiki/Space:United_States_Federal_Census

The main focus is to add everyone using personal scams of sheets,  but many are working from your thoughts,  as well.

Comments

That is very ambitious and in the long run will be very helpful if it can be accomplished. I have a feeling though before this project is complete, an easier solution will come along. This is the same approach I was using on a census I was doing. Later I got tired of linking everyone back to page. I think a template is a better idea in that I can just enter the template and the supporting structure is already there.

---

So, what I want to see as an end result, is a form table, like a free space page but not quite so primitive, that lists every single line entry on every single sheet in each census year.

And every name on that form will have a Wikilink alias to a matching profile. And that matching task will be automatically performed by applying a template on to the profile.

Is such a dream even possible in a manual or hybrid automated system? Or would it have to be fully automated, both on the form side, and on the profile side?

Answer 4

"It's gonna take time, a whole lot of precious time, it's gonna take patience and time, to do it right."

So I've got a George Harrison earworm (to share) as a result of this question.

I would love to be able to associate specific census pages - and identify lines on that page for specific ancestors.   I love the idea also of seeing who else is on the same page, or in the same household.

I'd wan't the ability to navigate through the census data in such a way as to travel household by household and page by page.

This requires a standard manner of reference and citation for each census.

(Tho I do note that while each record will point to a person, there are double counts.)

Future aims could include longitudinal navigation -  for a location or a family.  This has great potential and would be a boon to many researchers.

I don't see how categories would function and satisfy these design goals, and agree they would be cumbersome.

Templates- could work if designed in a succinct manner - and I think may be part of the solution - at least in so far as included on each profile.

I have been asking myself - how would I want to approach this as an open and collaborative information project like wikitree?   

Do we want to squeeze such efforts into the wikitree box and all that it entails for wikitree tech constraints, or do we want the basic utility for wikitree users, in conformity with wikitree's higher standards?

These ideas could only proceed with the blessing of Chris & team, but I think as a matter of long term functionality it has merit:

Either a new domain (TLD) under auspices of wikitree -- WikiCensus or CensusWiki (have not searched if taken) - with possible subdomains for each census year (could do for any and all censuses each country & state), or a subdomain of wikitree itself.

Data would thereby be sequestered in each year/census project, allowing slight variations for each page/template/nav (as the census tended to change certain elements/questions each year), though would follow similar pattern for navigation between pages, households and records.  

To some extent we have some of these features in familytree.org, though I wish it were in a Wiki allowing the community to correct and manage data in real time.

Summary for my imagined lowest cost approach ---  a dedicated wiki (perhaps as subdomain) with the wiki page unit being the census page data.  (I don't suppose census image is feasible, at least not for all censuses.  But at the bottom of each page we can link to multiple sources for images,  Indeed this has a benefit as different providers do have different images, and they vary in quality/legibility.)

Comments

In Swedish genealogy most people use a product Archive Digital that has unique identifiers for every page. 

The vision I have is to combine DNA genealogy with a query 

• For family tree members in people related to my DNA search result
• Find all people who have a source that is in the same church book as one person in my family tree inside 5 pages

Answer 5

Just a thought....

 

Don't Forget the UK Census

Comments

And birth registries, and city directories, and SSDI .... and... and... when oh when are we going to finally get an automated source-based genealogy linking platform? Please, make it so.

---

Not too bothered about things being automated. But a census system that lets you link profiles would be good.