Posted by Jennifer Holik on May 26, 2016 in Collections

Last month I wrote an article called Combining World War II Research on Fold3 with Your Ancestry Family Tree. This month we will look at Ancestry family trees in another way.

Ancestry family trees offer researchers a way to organize their research and share it with the world. We can search for a multitude of records and attach them to individuals on our tree, creating facts about their lives. One question we should ask ourselves is, “Are we attaching records correctly?”

Genealogy is a learning process for all of us, and many of us started by attaching everything, regardless of whether it was accurate, until our education developed further. Ancestry offers researchers information through this blog and Ancestry Academy. Ancestry Academy offers many courses to help new and experienced researchers learn about the records and help them create more accurate family trees. Understanding the records and how to properly attach them to our tree, benefits everyone.

I have seen many cases of World War II research where the facts are partially or completely incorrect, leading those browsing trees to assume, and often attach, incorrect information to their own trees. Using a few tips, researchers can better document their World War II ancestors in their family trees. Let me illustrate what I mean.

William F. Cowart
William F. Cowart

William F. Cowart 

William F. Cowart was a Marine during World War II. Born. 21 February 1922 at Reform, AL. Killed In Action on 20 November 1943 at Tarawa. Remains never recovered. Serial number 471443. Final unit: C Company 2ndAmpTrBn, 2ndMarDiv. (Source: USMC Casualty Card for William F. Cowart, serial no. 471443. Marine Corps History Division, Quantico, VA. Death Certificate for William F. Cowart, serial no. 471443. Individual Deceased Personnel File, AHRC Ft. Knox, KY.)

One family tree for William F. Cowart contains information attached about his military service, death, and burial which is somewhat inaccurate. There are ways to improve on this information by looking at the records more carefully.

Death Fact: The date of death listed is 21 November 1943. However, this date is incorrect due to indexing. If you examine the Headstone Application it has typed 21 November 1943 as the death date. But the military examined this record to ensure the facts were correct and crossed out 21 and wrote 20. According to his Individual Deceased Personnel File and USMC Casualty Card, he was Killed In Action the 20th of November. Cowart’s father completed or helped complete this Headstone Application. Families made errors all the time. Usually the government checked the information, but not always.

Fixing the Fact in the tree: The date should be changed to 20 November 1943. A notation could also be made in the Description Field of the fact to explain why the date is 20 and not 21 November. Unfortunately the death date cannot be changed in the index because the only change allowed is the year.

Burial Fact: The Burial is listed as Ethelsville, Pickens County, Alabama. Three sources “prove” this fact. Two Find A Grave entries and the U.S. Headstone Application. However, if the headstone/marker is examined when the researcher goes to FindAGrave, they will see it says “In Memory Of.” This is a memorial stone only. Cowart is not buried in Alabama.

Continuing the examination of Find A Grave entries for Cowart, we will also see several which were created by various sources, including the American Battle Monuments Commission (ABMC) for all those buried overseas or listed on a Wall or Tablet of the Missing, as Cowart is. One issue with Find A Grave, that I have no idea how to solve, is the multiple entries for Cowart make him look like he is his own sibling.

Researchers should examine all entries for Cowart because they provide different details. Together, they help paint a picture of Cowart’s life.

Further examining the Headstone Application, we see at the top it says NONRECOVERABLE. This means Cowart’s remains were never found.

U.S., Headstone Applications for Military Veterans, 1925-1963 for William Franklin Cowart
U.S., Headstone Applications for Military Veterans, 1925-1963 for William Franklin Cowart

Fixing the Burial Fact in the tree: Without the Individual Deceased Personnel File (IDPF) obtained from Ft. Knox, KY, Ancestry users will not know Cowart was actually identified and buried soon after the Battle of Tarawa, ON Tarawa. Unfortunately with the chaos of war, when the Marines left the island and the Navy Seabees arrived, they “beautified” the cemetery and many of the identified dead were now “lost.” Families received letters after the war stating their sons could not be found. Most are still “lost” today.

The burial fact can remain but the place of burial should not say Alabama. At the very least, until or if Cowart is recovered, it should say Tarawa or Gilbert Islands. If a researcher has the IDPF and can document a burial date, then it should be added.

A note in the description field of the fact could indicate Cowart’s remains were not recovered and he is not buried in Alabama. Only a memorial stone exists for him there. Records: Finally, offers researchers the ability to search for documents which help tell the story of their soldier. When we search for only William Cowart’s name on Fold3, a couple of entries appear, which we can also find on Ancestry. However, if we search within the WWII War Diaries for Tarawa, 2nd Marine, we find the Second Marine Division War Diary for November 1943.

World War II War Diaries, 1945. Record courtesy of Fold3.
World War II War Diaries, 1945. Record courtesy of Fold3.

This document likely does not list Cowart by name, but provides historical context to his story. These records may also be attached to individuals in our family trees.

Fixing Fold3 Records in the tree: Each page of a record can be attached individually and will create a new fact unless we download the document and then add it to a specific fact. If you only attach one page out of a full document, consider adding a note in the description field as to what this is and what other information it contains. You might also consider writing up a short document in Word summarizing the information, and attaching that to the fact.

We all have a lot to learn about our families, the records available, and how best to document this in our family trees. Being as accurate as possible helps to build more accurate trees and on some level, educate others how to do the research. It also further builds our relatives stories as we continue to honor their memories with our research.

Jennifer Holik

Jennifer Holik is an international WWII researcher, speaker, and author of the only authoritative books on the market, “Stories from the World War II Battlefield,” which teach individuals how to research WWII service across any branch. She can be found at her website The World War II Research and Writing Center or on Facebook.


  1. Great research problem, Jennifer! Thanks for walking us through this exercise. I also appreciated your honesty in reference to FAG, an Ancestry-owned property. FAG can be a great resource or a great headache, or even both. Still, it was helpful to see the process to determine these dates.

  2. Joyce

    I agree people do not pay attention to detail and it causes a lot of problems as then the folks who come behind just copy their mistakes. People depend on Find a Grave too much rather than verifying the info. People who make Find a Grave entries are just that –people–you have no idea how well they have researched something, or how careful they are in their research, yet people seem to think “if it is on Find a Grave it must be right”. I submit corrections to folks who do listings WITH SOURCES and most folks will change the listing…but not all. One in particular who does a LOT of New England records will not change anything, regardless of what proof you send her. Be careful with Find a Grave.

    I will be honest though, I get REALLY tired of fixing transcription errors on ancestry as it seems every time I turn around I am submitting a correction for something.

    What REALLY irks me though is that on many records the ONLY field we can correct is a name. I find problems in lots of fields and it seems to me that if a field is on the transcript we should be able to submit a correction.

    What I have been doing with those documents in submitting a report of the error through the “report issue” button but I doubt things will ever get fixed that way either.

    A few months ago I reported to HQ that nearly the entire database (transcriptions) of New England marriages before 1700 was incorrect as they transcribed the b. as a marriage date when the instructions for that database clearly say that the b. means the date of the first child born.

    Although HQ said they passed it along to “the powers that be” BUT also told me that it was a very low priority as it was ONLY a few people affected—That database (with supplements) is about 100,000 records.

    Most generations in that time period had 10-12 kids—do the math–that affects a LOT of people.

    I suggested that the only way to possibly fix these records was to change the page the transcript is on with a notation on it that the b. meant date of first child and with a link to the instructions on that database as Torrey has 2 pages of all kinds of things that enable you to interpret the records correctly.

    The people who transcribed the database were obviously not given the instruction OR did not understand them (there are so many different types of abbreviations he used I have to go back and check it every time I see one). Even the TYPE of parenthesis used means something…This is a most difficult database to use correctly and since b. means date of first child born, it seems to me that rather than correct every entry, it would be simpler to give info about the database on the transcript pages. Fixing it 100% is impossible. I suggested they just add, through programming, a short explanation about what the b. really means and to provide link to instructions and recommend people use that link to properly interpret the records. A simple fix that at least draws people’s attention to the fact that these transcripts need to be verified–SO many are wrong.

    Most people don’t bother to enter corrections. This hurts us all. BUT I am really tired of doing corrections especially for things that should not have to be corrected such as directories. I don’t understand why the folks who transcribed them could not read a typewritten document but it appears to me that they had no command of the English language as I have seen the craziest things on directories. I have even seen the old style phone numbers included as part of the address–and this is NOT just the occasional one. Yesterday i saw a directory for all people living on Lounsbury St and the person who transcribed it transcribed EVERY last name as Lounsbury–when a simple glance at what it said would have prevented the error. It was as plain as day at the top of about 30 people it said “Lounsbury Street” yet every one of those people was transcribed as Lounsbury even though the entries made NO sense the way thery were written.

    Ancestry really needs to do a better job of quality control on this stuff and when an issue is identified they should make every effort to correct it.

    We get tired of correcting ancestry transcription errors. I spend too much of my time submitting corrections!

    When transcription errors for entire databases or entire pages of a directory or time period (such as when the phone numbers used to be 2 letters and then numbers YET they were all transcribed as part of the address) ancestry should correct those things.

    Ancestry is creating problems by not fixing transcripts when they are reported to them!

  3. John Foltz

    Taking the entire ancestry list how can you fine duplicate at one time and it they are related.

  4. Janet

    Joyce, the city directories are OCR indexed. The computer reading the entries doesn’t have critical thinking skills.

  5. Sue

    Whenever I use Optical Reader software, I always have to go back and proofread the results. It’s irresponsible of Ancestry to use this time-saving device without doing so.

Comments are closed.