Wednesday, June 29, 2011

Citation Principles for Genealogy Record Publishers

The Insider's Guide to Citations

Citations have two purposes: locate the source and indicate its strength. This series of articles explains what we must do to accomplish these purposes for genealogical sources.

 

Yesterday I reviewed the fields displayed in the Social Security Death Index (SSDI) by Ancestry.com and FamilySearch.org. (See “SSDI: Ancestry.com vs. FamilySearch.org.”) Next week I would like to review their citations for this collection. Before I can do that, I have to present the criteria against which I will judge their citations.

To that end, here are my citation principles for record publishers.

  • Like professional genealogists, professional publishers of genealogical records must provide professional quality citations. Mills style is the standard used by most all of the professional community.
  • Publishers must provide citations for their record collections and for the individual records within their collections.
    • I’d better explain what I mean by “record collection” and “individual record.” According to Webster, a record is “something that recalls or relates past events; an official document that records the acts of a public body or officer; an authentic official copy of a document deposited with a legally designated officer.”1 For genealogy record publishing, I include a published derivative of a record.
    • A record collection is a titled group of records. The “Social Security Death Index” is a record collection. Ancestry.com uses the term database, which is the name of the technology used to implement a record collection.
    • Here are examples of user interface or record-level citations from various websites:
       Wikipedia cite this page tool Cite popup Cite article popup Cite this page, hover help Citation text insertion
  • Citations to published collections differ from citations to the sources of the collections. The two should not be equated or mislabeled.
  • Citations to published collections should include source-of-the-source citations to the sources of the collections.
  • Citations to records must contain the information necessary for users to locate the published records, notwithstanding website changes. 
  • Citations to records must also enable location of archive originals. If records from multiple archives are published as a collection, each record citation must specify the source archive.
  • Source list (bibliography) citations differ from reference note citations. Publishers may wish to label citations appropriately.

Next time I’ll review the citations provided by Ancestry.com and FamilySearch with their SSDI collections.


Sources

     1.  Merriam-Webster Online Dictionary, online edition (www.m-w.com : accessed 23 November 2009), “record.”

Tuesday, June 28, 2011

SSDI: Ancestry.com vs. FamilySearch.org

I mentioned last time some of the shortcomings of the Social Security Index (SSDI). See How Reliable is the SSDI?

Consider the “Verified or Proven” code supplied by the Social Security Administration in the public Death Master File. The code indicates that a record in the SSDI has been verified or a death certificate has proven the death information. Neither Ancestry.com nor FamilySearch.org display this

Example SSDI record from Ancestry.com
Example SSDI record from Ancestry.com

 

 Example SSDI record from FamilySearch.org
Example SSDI record from FamilySearch.org

This table compares the fields from the Social Security Administration’s Death Master File1 displayed by Ancestry.com and FamilySearch.org.

Field Ancestry.com FamilySearch.org
1. Change status: Record has been added, changed, or deleted Not applicable Not applicable
2. Social Security Number Displayed
and used to derive the “State [and] (Year) SSN issued”
Displayed
and used to derive the “place of issuance”
3. Last Name Concatenated fields 3-6 Listed separately
4. Name Suffix Concatenated fields 3-6 Listed separately
5. First Name Concatenated fields 3-6 Listed separately
6. Middle Name Concatenated fields 3-6 Listed separately
7. Verify or Proof code Not displayed Not displayed
8. Date of Death (in MMDDYYYY format) Ancestry.com reformatted the date to genealogical standard, with month abbreviated. FamilySearch reformatted the date to genealogical standard, month not abbreviated.
9. Date of Birth (in MMDDYYYY format) Same as death date Same as death date
and used to calculate
“estimated age at death”
10. State/Country of Residence (for deaths reported before 2/88) I was not able to tell if either vendor uses this information I was not able to tell if either vendor uses this information
11. Zip code, Last Residence Displayed
and used to calculate city, state, country
Displayed
and used to calculate city, state
12. Zip code, Lump Sum Payment Displayed as “Last Benefit”
and used to calculate city, state, country
Not displayed
13. Citation. The source of each record is not specified in the DMF, so websites can not give record level source-of-the source citations. Displayed record-level and collection level citations Not displayed

Commentary

2. It is nice that Ancestry.com displayed information for both when and where the number was issued. This may help place the person at a particular time and place. Remember that issuance date and state is derived and is subject to error. And remember that the state is not necessarily where the recipient resided.

FamilySearch displayed the field names in all lowercase. There are two established standards (title style and sentence style) for capitalizing titles in the English language and FamilySearch deviates from both. Is there a good reason for that? I would have thought that mixed case would be easier to comprehend. (See my “Indexing Errors: Test, Check the Boxes.”)

3 through 6. There are instances where websites should not manipulate information before display; they always seem to screw up one or more records. I think name concatenation is pretty safe; I like it because I can comprehend the names on Ancestry.com much easier than FamilySearch’s non-concatenated names.

7. Failure to display the Verify or Proof Code is a major flaw in both websites. Even though more than 90% of the deaths are not verified, for times that there is verification or proof of death, that is important to know. If you want to see this field, you can use RootsWeb.com.

8. Ancestry.com abbreviated month names. I know Mills citation style recommends spellings them out in citations. Does BCG have a recommend? In columnar lists of results, Ancestry.com utilizes the screen real estate better than FamilySearch, in part by using abbreviations. (See below.)

image image
I prefer Ancestry.com’s columnar lists over FamilySearch’s.
FamilySearch wastes a lot of space.

9. When a birth date is present, FamilySearch estimates the age at death. That’s a great practice. I predict you’ll see Ancestry.com add this same feature once today’s article alerts them of the possibility.

11. Both Ancestry.com and FamilySearch use the field title supplied in the DMF, but I question the title of this field. According to other sources this field is the last address on record. I’m guessing that it is actually the last mailing address associated with the account. I get an annual statement from the SSA; don’t you? I’m guessing it is often not the last residence. It could be a post office box, or the address of a relative or legal guardian.

12. Would someone tell me what a “lump sum payment” is? I’m guessing that monthly SSI or retirement benefits are not lump sum benefits. If that is the case, then Ancestry.com has mislabeled this field.

For both 11 and 12, keep in mind that the only information supplied in the DMF is the zip code. The zip codes in the DMF are not changed as zip code assignments change over time, so these websites might be showing incorrect information. Take the example of Donald N. Sider, shown in the list results above. (I wonder if we’re related!? My side of the family changed our name from “N’Sider” to “Insider” when we immigrated to Fantasyland from Frontierland.)

Donald N. Sider’s last residence was zip code 33413. According to the U.S. Postal Service, 33413 currently applies to both West Palm Beach and Greenacres, Florida. Ancestry.com displayed one of the two, West Palm Beach. FamilySearch did not display either city, but did display the county.

13. I’ll review the citations later in a separate article. Stay tuned…


Sources

     1.  National Technical Information Service, “Death Master File Record Format,” NTIS [Death Master File website] (https://dmf.ntis.gov : accessed 15 June 2011), select “Record Layout.”

Friday, June 24, 2011

Both Male and Female?

Records Say the Darnedest ThingsRecords say the darnedest things

We depend upon records to reveal the “truth” about our pasts.

Yet sometimes records have anomalies.
Some are amusing or humorous.
Some are interesting or weird.
Some are peculiar or suspicious.
Some are infuriating, even downright laughable.

Yes, Records Say the Darnedest Things.”

Both Male and Female?

There are only two genders, right? Well… Maybe…

I came across this record recently while indexing Maine vital records:

Gender both male and female

As it turned out, the card documented the stillbirth of twins.

Stillborn twins

I was grateful that Indexer A knew what she was doing. She had added a second record to the image. One record documented the boy, a second documented the girl.

Yes, “Records Say the Darnedest Things.”

Tuesday, June 21, 2011

How Reliable is the SSDI?

How good is the information from the Social Security Death Index (SSDI)? It may not be as good as you think.

“The full DMF [Death Master File] includes both the verified and unverified reports of death for Social Security beneficiaries and non-beneficiaries,” said Bill Gray, an official of the Social Security Administration (SSA). The DMF is the government’s name for the Social Security Death Index. “We receive approximately 2.5 million death reports each year from many sources. We receive 90 percent of the reports from family members and funeral homes, with the remainder coming from States and other Federal agencies through data exchanges and reports from postal authorities and financial institutions.”1

According to Gray, over 90% of the information is unverified. Because studies showed that the Information obtained from family members and funeral homes is 99% accurate, SSA does not verify it. Information obtained about individuals who were not receiving Social Security benefits is also not verified, as the agency does not have the necessary contact information. Otherwise, deaths are verified through family members, payees, or medical institutions.

Gray said that states are allowed to prohibit redisclosure of death information they provide. The SSA removes such unverified state information to create the public version of the DMF. Genealogy websites rename and publish the DMF as the Social Security Death Index. Many supplement the DMF with additional information. Some websites add the state of issue, deriving it from the numbers in the SSN. Some provide city and state for zip codes. Users are advised to consider the city name with care, as the DMF is not updated as zip code boundaries change. Further, zip codes may cover multiple cities and genealogy websites may not display the correct one.

Gray considers the data to be 99.5% accurate. But as you evaluate the strength of the information, keep the following in mind.

  • The SSDI does not list the source of each entry nor does it indicate how the information was verified.
  • Most websites do not show the “Verify or Proof Code” from the DMF. This code indicates entries that have been verified and entries that have been proven by death certificate.
  • If the verified code is not displayed and the person died before retirement age, chances are he or she were not receiving benefits, and thus the information was not verified.

The SSDI does not contain information on every death. While coverage has increased dramatically since the early 1960s, coverage for the young remains low because they die before receiving benefits.2

SSDI coverage varies by age 
Estimated percentage of deaths included in the SSDI, by age, 1960-1997
Image credit: Hill and Rosenwaike

Next week: The SSDI on Ancestry.com and FamilySearch.org. 


Sources

     1.  Bill Gray, “Testimony before the Permanent Subcommittee on Investigations of the Senate Committee on Homeland Security,” 9 July 2008, Social Security Online (http://www.socialsecurity.gov : accessed 14 June 2011), legislation and congressional affairs.

     2.  Mark E. Hill and Ira Rosenwaike, “The Social Security Administration’s Death Master File: The Completeness of Death Reporting at Older Ages,” Social Security Bulletin 64, no. 1 (2001/2002): 48; PDF online (http://www.socialsecurity.gov/policy/docs/ssb/v64n1/v64n1p45.pdf : accessed 14 June 2011), “Research Statistics, and Policy Analysis” > “Research and Analysis.”

     3.  “Social Security Testimony Before Congress : Ways and Means Social Security Subcommittee,” 8 November 2001, Social Security Online (http://www.ssa.gov : accessed 14 June 2011), legislation and congressional affairs.

Friday, June 17, 2011

Darned Quarrels

Records Say the Darnedest Things

We depend upon records to reveal the “truth” about our pasts.

Yet sometimes records have anomalies.
Some are amusing or humorous.
Some are interesting or weird.
Some are peculiar or suspicious.
Some are infuriating, even downright laughable.

Yes, Records Say the Darnedest Things.”

Darned Quarrels

As we try to understand older records, it is important to remember that the meanings of words shift over time. If your ancestor’s occupation was “quarrel-picker,” you might mistakenly conclude that he was a fighter. Obviously, a “quarrel-picker” is someone who picks quarrels.1

To pick a quarrel

But “quarrel” is a homonym. It can also refer to a quadrilateral shaped object. At one point in history a a square or diamond shaped piece of glass was called a quarrel. Your quarrel picking ancestor was nothing but a harmless glazier.2

Well, I suppose your ancestor was harmless… unless he was a shard wielding, quarrel-picking quarrel picker…

Yes, records say the darnedest things.


Sources

     1.  Abel Boyer, The Royal Dictionary Abridged in Two Parts (London: 1728), s.v. “pick;” images online, Google Books (http://books.google.com : accessed 6 June 2011), id=4ChFAAAAcAAJ.

     2.  “Diogenes Robb’d of His Tub: or, the World Despis’d,” State Tracts: Containing Many Observations and Reflections, ed. Joseph Browne and William Oldisworth (London: 1715), 90; images online, Google Books (http://books.google.com : accessed 6 June 2011), id=BkFWAAAAYAAJ. This tract was perhaps from the periodical Growler, or Diogenes Robbed of his Tub, published 1 February 1710/11, perhaps by Abel Boyer.

     Jeffrey Kacirk’s Forgotten English daily calendar provided this example.

Wednesday, June 15, 2011

Emphasize the Source You Used Over the Source of the Source

 Two derivatives of the 1790 Census, Brookfield, Connecticut
Above are two types of derivatives from
the same page of the 1790 U.S. Census.

Citations have two purposes: locate the source and indicate its strength. This series of articles explains what we must do to accomplish these purposes for genealogical sources.

Last time (see “Citing Quoted Sources.”) we learned how Chicago handles citations to information that is copied (“quoted”) from another source.

Mills swaps the order of the two citations, and replaces “quoted in” with “citing” (or another term that makes sense). This gives emphasis to the source seen by the researcher. I like that; it makes more sense to me.

     3. Milton Rubincam, Pitfalls in Genealogical Research (Salt Lake City: Ancestry, 1987), 11; citing Donald Lines Jacobus, “Tradition and Family History,” The American Genealogist, 9:1 (July 1932).

The first row in this table shows Chicago. The remaining rows show examples from Mills.

First citation   Second citation Examples in Mills’s Evidence Explained
Citation to the source of the source quoted in Citation to the source used by the researcher Chicago
Citation to the source used by the researcher citing Citation to the source of the source pp. 166, 348, 404, 427, 577, 630, 786
Citation to the source used by the researcher from Citation to the source of the source pp. 166, 577
Citation to the source used by the researcher crediting Citation to the source of the source pp. 94, 128

 

The examples on pages 94 and 166 are interesting because the citation to the source of the source is given inside quotation marks, showing it was copied exactly as it appeared in the source. (Also see page 446 for a different way to cite the source of the source.)

In closing, it bears repeating that when the source you use quotes another source, you need to get that source and use it if at all possible. “Never drink downstream from the cows.”1 And never, ever cite a source that you did not see.2


Sources

     1.  Ancestry Insider, “Never Drink Downstream from the Cows,” The Ancestry Insider (http://ancestryinsider.blogspot.com : 25 February 2011, accessed 9 June 2011).

     2.  Elizabeth Shown Mills, Evidence Explained: Citing History Sources from Artifacts to Cyberspace, PDF images (Baltimore: Genealogical Publishing Company, 2007), 52.

Tuesday, June 14, 2011

101 Best Websites

Ancestry Insider is one of the 101 Best Websites 2011Family Tree Magazine has announced their list of the “101 Best Websites for 2011” and has chosen again to honor the Ancestry Insider. With the growing number of awesome genealogy websites, it is a growing honor. I’m humbled. With the growing demands of my day job, it is a growing challenge to find the time. I have a growing worry that I’ll fail to provide enough value to make it worth your time to follow my ramblings.

I suppose that is one reason why I’ve migrated from Family Tree Magazine’s news category into their “Tech Tools” category. I can no longer spend the time necessary for a good news blog. But technology is my forte.

“With all the updates at FamilySearch, it’s more of a must-read than ever,” wrote David A. Fryxell of the Ancestry Insider.

Some of the other winners in the tech category are Google book and newspaper archives, Internet Archive, and genealogy search engines: Mocavo, Live Roots, and Steve Morse’s One-Step. Some of the other categories include Family History Mega-Marts (websites with the largest record collections) and State-of-the-Art Archives (state archives).

See “101 Best Websites for 2011” to see all this year’s winners with Fryxell’s commentary.

Ancestry Insider is one of the 101 Best Web Sites 2008Ancestry Insider is one of the 101 Best Web Sites 2009Ancestry Insider is one of the 101 Best Websites 2010

Friday, June 10, 2011

Soil? Inconceivable!

Records say the darnedest things

We depend upon records to reveal the “truth” about our pasts.

Yet sometimes records have anomalies.
Some are amusing or humorous.
Some are interesting or weird.
Some are peculiar or suspicious.
Some are infuriating, even downright laughable.

Yes, Records Say the Darnedest Things.”

Records Say the Darnedest Things: Soil? Inconceivable!

Sometimes words don’t mean what you think they mean. Keep that in mind when reading old records. I find it especially interesting when words turn a full 180.

Consider “soil.” According to Webster it means to stain, defile, blacken, or taint.1 But when it comes to milk, it means the opposite.2

Soil can mean to defile or to clean
“To Soil milk, to cleanse it”

Here’s a weird one. Cover your eyes if you blush easy. Don’t misunderstand pre-reformation wedding vows. When the wife pledged to be “buxom in bed and at board” she was only promising obedience.3

Don't misunderstand old wedding vows

Yes, records say the darnedest things.


Sources

     1.  Merriam-Webster Online Dictionary (http://www.m-w.com : accessed 4 June 2011), s.v. “soil.”

     2. John Ray, A Collection of English Words (London: H. Bruges, 1674), 44; images online, Google Books (http://books.google.com : accessed 4 June 2011), id=njdWAAAAYAAJ.

     3.  Samuel Johnson, A Dictionary of the English Language in Which the Words are Deduced from Their Originals, 9th ed., vol. 1 (London: 1805), s.v. “buxom;” images online, Google Books (http://books.google.com : accessed 4 June 2011), id=RaQRAAAAIAAJ.

     I acknowledge Jeffrey Kacirk’s Forgotten English daily calendar for these examples.

Wednesday, June 8, 2011

Citing Quoted Sources

The Insider's Guide to Citations

Citations have two purposes: locate the source and indicate its strength. This series of articles explains what we must do to accomplish these purposes for genealogical sources.

 

Citing Quoted Sources

Remember the big no-no in writing school papers? You’re reading a book and it quotes another. You like the quote and use it in your paper. You’re supposed to go find the original and cite it. If you can’t, you must cite the first book and its citation of the original. The Chicago Manual of Style says such a citation should look like this:1

     2.  Donald Lines Jacobus, “Tradition and Family History,” The American Genealogist, 9:1 (July 1932); quoted in Milton Rubincam, Pitfalls in Genealogical Research (Salt Lake City: Ancestry, 1987), 11.

Do you see how this is two citations in one? Remember the form; we will use it for derivative sources.

Citation of the source of the source ; quoted in Citation of the source seen by the researcher

 

I use a semicolon to separate the two, since that is the English language standard for separating list items that contain internal commas.2


Sources

     1.  The Chicago Manual of Style 15th ed. (Chicago: The University of Chicago Press, 2003; CD-ROM version 1.2.3), 727.

     2.  Elizabeth Shown Mills, Evidence Explained: Citing History Sources from Artifacts to Cyberspace, PDF images (Baltimore: Genealogical Publishing Company, 2007), 88.

Tuesday, June 7, 2011

Stop and Wait Before You Update

Once again there were more problems with FamilySearch Indexing over the weekend. Once again the indexing servers seemed to have crashed. Once again users saw this warning:

image

If you get this message when submitting a batch, do not update your FamilySearch Indexing software without first learning if the indexing server is up or down. Wait a couple of minutes and try again to submit the batch. If you still cannot submit, exit the indexing software and restart it. If the indexing server is down, the indexing software will not restart. Instead, you will see this:

image

This is your signal to not update. Instead, wait minutes, hours, overnight even, and then try again.

FamilySearch issued a statement during the outage Sunday advising indexers that Sunday afternoon outages may continue for a time. Indexers were encouraged to download extra batches beforehand.

Here is the complete text of the statement.

We understand that our indexing system still struggles a bit at our busiest times of the week—usually Sunday evening or Monday morning depending on your time zone. If you are concerned about running into problems during this time, we have some steps you may want to follow. You do not need to do this, but it may help avoid possible issues.
  1. Before 1:00 p.m. on Sunday (MDT, UTC-6), download up to 10 batches.
  2. Open each batch to make sure the information is saved to your computer.
  3. On the indexing start page (or the My Work page), click the Work Offline box.
  4. Click OK. Note: You do not need to disconnect your computer from the Internet; you are simply telling the program not to connect to the indexing server while you index these batches.
  5. Do not close your indexing program. Keep your computer on and your indexing program open until you have completed and submitted all 10 batches.
  6. Work on your batches. You will not be able to submit or save your batches while in "offline" mode.
  7. When you have completed all of your batches, return to the indexing start page, and click the Work Offline box to remove the checkmark and return to "online" mode.
  8. Click OK on the dialog box. Note: You should already be connected to the Internet.
  9. Highlight each batch, and click Submit Batch.
  10. Download more batches if you like, and start over at step 2.

If, while submitting your batches, you receive a message to update your program, this is an error. You do not need to update. Simply click OK. Let the program close. Wait a few minutes, and open the program again using the desktop icon. Once the program is open, try submitting your batch again.

Thank you for your patience during these busy times. We hope to get the system to a point soon where the program can handle these busy times smoothly.

Friday, June 3, 2011

Search Ancestry.com with Search Plugin

One reason you're not hearing much from me is the crash of my computer. Fortunately, I've followed the advice of people like Dick Eastman and have what I hope is a complete backup of my documents. But buying a new computer and getting all my programs reinstalled is taking a lot of time, time that is already overly busy for various reasons.

As I tried to figure out a new version of Firefox, I was reminded that it is possible to add additional search engines to the Search Box, located near the upper right corner of the browser. I wondered how easy it would be to create an Ancestry.com Search Plugin.

It was easy. The problem is, it is nearly worthless. I've mentioned before that I think a Google-type search--a single text box--would not work well for genealogy records. This experiment has done nothing to change that opinion.

It is an interesting toy to play with. To install the plugin, go to this page and click on Ancestry.com. Type in names and dates and places. Try someone famous, like "john fitzgerald kennedy." I'm certain that if Ancestry.com put their minds to it, they might vastly improve how well a single-text field search works. But I think tens of man years would never make it so usable that you would choose a single text field over separate name, date, and place fields.

I created this Firefox search plugin at http://mycroft.mozdev.org/submitos.html

This has been a nice diversion, but it is time to get back to setting up my new computer. Stay tuned...