Ancestry.com

U.S. Content Update: Ancestry-FamilySearch merged 1900 U.S. Census index now live…and FREE

Posted by Ancestry.com on August 29, 2008 in Content

Just in time for FGS, the new and improved 1900 U.S. Federal Census is now live and free for a limited time, the first product of our U.S. Census agreement with FamilySearch.  So what’s new and improved about it?  Well, here are some basics:

  • NEW IMAGES: Ancestry’s images, produced several years ago, were replaced with FamilySearch images, created recently from earlier generation microfilm and with the latest imaging technology.  In the images below, a piece of tape that obscured text in the census image on Ancestry (top) is more transparent on the recently added FamilySearch image (bottom):

    ancestry_tape.jpg

    fs_tape.jpg

  • NEW INDEX: We took our existing index with user-submitted corrections, and the new, double-keyed and arbitrated FamilySearch index and programmatically merged the two.  A complex algorithm was designed specifically for this operation.  The following are a couple quotes from one of our developers:
    • “If a TGN (Ancestry) record was not linked [to a FamilySearch record], we also did a check to see if we could identify the FS (FamilySearch) image.  If we got a match on state, county, township, ED, page number . . . then we would use the higher quality FS image in place of the TGN image. In summary, where possible a FS image was used in place of the TGN image.”
    • “Also, FS provided more fielded data (several fields) in addition to what we had . . .  Merging the two gives a more complete keyed set. Example: Where we estimated the birth year, they keyed it. Where they estimated the age, we had that keyed.”
  • REMAINING FIXES: We do know of some issues with the browse, as well as some other items, that need to be fixed and we’re already working on those.  If you find bugs or errors, please forward them on to me.

Again, for a limited time, the index is free.  Search and view the data to your heart’s content.  I hope you’ll take advantage of the first of the new and improved U.S. Federal Census databases on Ancestry.com.

15 comments

Comments
1 NickAugust 29, 2008 at 11:02 am

Keep up the good work

2 Charles A. Caul, SrAugust 29, 2008 at 9:40 pm

How do I “Select one parent in order to add new sibling” when NONE are shown?

3 PhillipAugust 30, 2008 at 6:31 am

The new index for the 1900 Census is broken. As an example, search for “George”, no last name, in Jefferson County, Tennessee, born in 1872 +/- 2 years. Not only are all Georges listed, some spouses, children, and even sisters of someone named George are listed. At least one entry, for “Jinnie Riley”, doesn’t include a George at all. Am I misunderstanding something?

4 RobertAugust 30, 2008 at 7:29 am

To Chris Lydiksen The new, improved 1900 U.S. census is a BIG JOKE was more poorly transcribed than the old one. If a person was born in INDIAN TERRITORY, some uneducated transcriber has them born in INDIANA, so to the volunteers that did the 1900 U.S. census If you can not cook get OUT of the kitchen, and to the persons in charge of this project are the most UNEDUCATED persons employed by ANCESTRY to let some thing like this go by unchecked, they have NO PRIDE in what they are doing. If I were the person in charged of Ancestry I would Fire all of them!

5 Zadruga GuyAugust 30, 2008 at 10:22 am

Robert, the transcriptions for the new version of 1900 census were done by volunteers at FamilySearch Indexing. So Ancestry.com had nothing to do with the transcription mistakes that are in the new version. FamilySearch Indexing does no checking for errors when both of its volunteers key in the same thing. So I guess this goes to show that no method of census indexing is error free.

6 Carol A. H.August 30, 2008 at 4:43 pm

I’m wondering about the corrections I made to so many people. According to this, they are gone:

NEW INDEX: We took our existing index with user-submitted corrections, and the new, double-keyed and arbitrated FamilySearch index and programmatically merged the two. A complex algorithm was designed specifically for this operation. The following are a couple quotes from one of our developers:

I guess I should go to the “new improved” 1900 census index and check for all the corrections I made.

7 Carol A. H.August 30, 2008 at 5:10 pm

I checked one family for whom I had entered customer name corrections and the corrections were still there. Phew!!! Hope all of them are there. I saved all my “thank you” emails from Ancestry so this should be easy to do. Of course the old index entry is still there. I didn’t expect that to change.

8 PamAugust 30, 2008 at 7:37 pm

To Robert:

I have indexed for FamilySearch and it isn’t easy. When you download a census page to index, you select the state and the county is randomly sent to you. I have indexed census pages in Ohio, Texas, Louisiana, and Florida. I had no knowledge of any of the family names in the counties. Sometimes the enumerator’s handwriting is very hard to read and after indexing 40 people per page on the 1870 census, you began to get tired and your eyes are strained. Many times I have had to make a guess as to what the names were and where the people were born. If you are unhappy with the indexing, I recommend that you volunteer to index and maybe you can help to improve the situation.

9 RobertAugust 31, 2008 at 9:50 am

To Pam

I volunteer to index to see what is was all about and got out of the kitchen, like you said you had no knowledge of the area you were indexing, or any background knowledge of handwriting in that period of time. We paid good money to Ancestry for their services, but lately we are not receiving the quality of there databases, like the North Carolina death certificates 1901-1975 If you are looking for some one that died in SHOOTING CREEK, CLAY,NORTH CAROLINA Ancestry has over 20 difference spellings of SHOOTING CREEK like SHORTING CREEK
SHOATING CREEK
SMOKY CREEK
SHODBING CREEK
SHOOTREY CREEK
SHERLING CREEK and the best ones SHVOTIRIGCRECH
SHROTUZ CRUTZ
SHWTING CRUK
Like I said before the person in charge of this database should be fired, you would think that they would give you a index of the spellings of the places in that state, but Ancestry wanted to be the first to get it out even if it is full of mistakes they have NO PRIDE in what they are doing
When the family trees started I thought that Ancestry would fix there index with the records attached to a person in your tree, I have 20696 persons in my tree and 52808 records attached
And to Ancestry stop wasting time and money in trying to come up with a NEW SEARCH ENGINE that will NOT work on the databases you have FIX the databases, Hire qualified people to do the indexing not some volunteer that has no knowledge of what they are doing, be the first to have the most correct index not the largest
Work on the user-submitted corrections to add corrections on not just the name but other fields like age, gender,relationship
It makes one wonder how many people that work for Ancestry has a family tree on Ancestry with records attached? and have first hand experience of trying to find some one in Ancestry databases

10 Andy HatchettAugust 31, 2008 at 12:37 pm

Pam, yopu said…

[quote]
I have indexed for FamilySearch and it isn’t easy. When you download a census page to index, you select the state and the county is randomly sent to you. I have indexed census pages in Ohio, Texas, Louisiana, and Florida. I had no knowledge of any of the family names in the counties. Sometimes the enumerator’s handwriting is very hard to read and after indexing 40 people per page on the 1870 census, you began to get tired and your eyes are strained. Many times I have had to make a guess as to what the names were and where the people were born. If you are unhappy with the indexing, I recommend that you volunteer to index and maybe you can help to improve the situation.
[End quote]

I’ll be blunt-

If you are having to guess then you are doing it wrong. Simply mark it as unreadable and move on.

Your incorrect guess plus another indexer’s matching incorrect guess leads to an incorrect index.

This goes for anyone indexing anything for any organization-period.

Indexers are there to help with a solution and not to compound the problem.

And for those who ask…
Yes, I index for both LDS and Ancestry.

11 RobertAugust 31, 2008 at 6:38 pm

TO CHIRS LYDIKSEN 1900 Census

REMAINING FIXES: We do know of some issues with the browse, as well as some other items, that need to be fixed and we’re already working on those. If you find bugs or errors, please forward them on to me.

when searching 1900 census Indian Territory, birthplace Georgia, birth year 1858 + _ 10 NO MATCHES FOUND take out the birth year and you have 8,359 found and not one person index has a birth date listed. like I said before be the first to get it out even if it is full of mistakes. If you were a employee of mine you would be FIRED!!! you did not impress me.
how can Ancestry NEW SEARCH ENGINE work right if the info is not in the database so lets blame it on some technical problem to cover your (you know what) IF YOU DON’T HAVE TIME TO DO IF RIGHT THE FIRST TIME, HOW ARE YOU GOING TO FIND TIME TO DO TO MAKE IT RIGHT

12 gloria baxterSeptember 1, 2008 at 6:09 am

I received the e-mail last night saying that this census would be free for limited time. When I went there it said ‘hope you enjoyed your 3 free days’!!! This is second time I have had this experience. Might be nice to receive the notice BEFORE it ends!

13 Yolande CARTERSeptember 1, 2008 at 10:38 am

My great uncle John Ratajczak born in 1866 is accounted for on the 1910 Cleveland Census. Immigration date 1895, married to Mary with two children Casimir and Stanislaw. I have been unable to locate him in the 1900 Census nor his children, in the 1920 and 1930 censuses. 1910 is my only reference to him and I am in a dead-end situation. Thank you for the 1900 census. Thought I finally could trace him but no luck!

14 JudySeptember 1, 2008 at 8:59 pm

I am absolutely thrilled with this newly improved 1900 Census. Prior to this release I had been unsuccessfully looking for several of my husband’s ancestors’ data for a year. There was a HUGE time gap, with many missing pieces to his ancestral puzzle. Yesterday when I searched the newly revised data it was like hitting the jackpot! Not only did I find the missing Census data, it led me to accidentally discover parents & siblings. I was able to push through a wall & find my husband’s great-great-grandparents. This discovery led me to marriages & children of people we didn’t know about, which sadly led me to a horrific discovery of how my husband’s great-great grandmother (who we’d just learned about) was brutally murdered, and her daughter nearly so. Yikes!

So anyways, thank you for this re-release of the Census data. For our family, at least, it worked wonders! I am eager to start searching it again to fill in a few gaps in my own ancestral line.

15 Chris LydiksenSeptember 1, 2008 at 9:16 pm

Responding to relevant comments:

3) Phillip, was that an exact search? New search interface, or old?

4) Robert, for certain, there are transcription errors, that goes with the territory where billions of characters are keyed into a computer. That said, the accuracy of the current index is higher than the index it is replacing. Fixing each and every transcription error and/or normalizing all relevant fields in a database would increase the cost and time of each database by a hundredfold or more, which would be less optimal for all involved, the business and its customers. We could simply just not key as many fields as we do, but the majority of entries are transcribed correctly. This adds value to the database by allowing users to search instead of browse.

6) Carol, user corrections were not deleted, but included in the merge. If you do find specific examples that contradict this, please let me know.

9) Robert, along with what I’ve already said above, the “Shooting Creek” value was not in a field that was normalized, rather it was keyed as seen and allowed to remain as such. Personally, I would love to see 100% accuracy, just like I’d love to fly on airplanes that are 100% safe, but I don’t want to pay $1 million per ticket. From the glass half full perspective, though, our search technology is constantly being improved to handle misspellings more intelligently. As for me needing to be fired and having no pride in my work, I can assure you at least the latter is untrue. I would love to sit down and discuss these issues with you man to man, and not via anonymous blog firebombs. Will you be attending any conferences soon?

12) Robert, was that an exact search? New search interface, or old?

13) Gloria, the 1900 index is free but you do need to be registered and logged in. If you still cannot access it, please call support.

14) Yolande, I’m sorry you are having difficulty finding your great uncle. The best advice I can give is to keep trying new searches, different spellings. Keep tinkering, but if search isn’t doing to for you, try browsing known residence locations.

About the Ancestry.com blog

Here you will find informational, and sometimes fun, posts from the folks behind the scenes here at Ancestry.com. We hope you’ll notice just how passionate we are about family history and about the products we’re building to help connect families over distance and time.

Visit Ancestry.com
Notifications

Receive updates from the Ancestry.com blog Learn more