Choosing and Using the Most Reliable Sources

I'm so close to finishing the source citation clean-up project I started in January. It's been massive. My tree has more than 57,000 people, so I have tons of citations. And this review doesn't include my 1,000s of Italian vital record citations. I'll get to them next.

This review involves making sure each citation has:

• facts about the source
• specifics of the citation
• a link to the record, and
• an image if available.

Also, importantly, I'm making sure citations get shared, not duplicated. For example, bad Ancestry syncs caused one census citation for a family of 8 to split into 8 source citations. It was mayhem. That ends now.

I went through my source list alphabetically. But I saved a few big ones for last. They included the:

• Social Security Death Index (SSDI)
• Social Security Applications and Claims Index
• U.S. Public Records Indexes, volumes 1 and 2
• U.S. City Directories
• Find a Grave

For my U.S. Public Records Index source citations, my goal was to delete each one where I had a better source. Why? Because too many times I've seen this source say that a person's birthday was the 1st of the month when I knew better. These were estimates. They were definitely not reliable. So, if I had a reliable source for someone's birthday, I deleted their Public Records Index citation.

When you have a handful of sources telling you the same facts, do you need them all?

The Most Reliable Sources

Most of my family lived in New York City until the last several decades. That means I have access to many of their birth, marriage, and death records online. These are big clear images of the documents themselves. They haven't been available online for long, so I'm still finding and downloading more and more.

Actual vital record images are very reliable for dates and places. I know my maternal grandmother was born on 25 Oct 1899 even though her grave marker purposely says 1900. Names are subject to spelling errors and variations, but if you keep an open mind, you can gather what you need.

If I have an image of a relative's vital record, any other source that happens to be correct is a supporting source. And some supporting sources are more respected than others. I'm putting my faith in the birth record written on the day. If the SSDI happens to have the same date, it's a nice supporting source. But if a Public Records Index happens to be right too, do I need to cite it? Not if I have the real thing. It's not a reliable source, and it adds nothing.

U.S. WWI and WWII draft registration cards are well respected sources, too. Of course your relatives born in the late 1800s may not be 100% certain about their birth date. So mistakes can happen. But if you've got a young man born in the 1920s registering for the WWII draft, I'd bet he knows his birth date and won't make a mistake.

Post-1890s ship manifests can be reliable, especially for identifying hometowns. Naturalization papers often have a lot of correct dates and places. These are both great to have when they support your most reliable sources.

Make sure you're choosing the most reliable sources for your family tree facts.

Less Reliable Sources

My own birth is on Ancestry.com as an image of a New York City index page. It's not a very clear image, so the certificate number is open to interpretation. And I didn't have a first name on day one, so how could anyone be sure Female Iamarino is me?

Last year I ordered my paternal grandmother's upstate New York birth certificate. I hope I got the certificate number right. She's listed as Female Merino instead of Lucy Iamarino, which is why I couldn't find her for so long. I'm eager to get the certificate to finally confirm her birth date. If it's really the 10th, my dad had it wrong his whole life. And my son was born on her birthday, but we never knew it! My son is 30.

When you have a bad index image or a database of facts pulled from indexes, it can't compare to seeing the original document.

The 1900 U.S. census includes the month and year of each person's birth. How nice! But that's only as reliable as the person who talked to the census taker. A census is a reliable source for home addresses, but not much else. The censuses are fantastic supporting documents unless they have a glaring error.

When I went through my U.S. Public Records Index citations, I planned to keep them only if they were my only source for a name, birth date, or death date. I whittled the citations down from 132 to 37. Most of what I kept is the only source I can find for a contemporary person's date of birth.

The two domestic sources I have left to review are U.S. City Directories and Find a Grave. I love when I find a grave marker image or an obituary on Find a Grave, but I know they're not reliable. My grandmother's grave marker says she was born in 1900 because she hated that she was born in 1899. My aunt told me she requested it that way because otherwise "my mother would kill me." And an obituary writer may not have the facts straight.

I'm sure I'll keep every Find a Grave source citation because they're often the only way to know where someone is buried. But if they disagree with a birth or death date from a reliable source, I won't attach the Find a Grave citation to that fact.

You may not want to launch a months-long review of all your citations, so here's what you can do. Each time you're working on a particular family unit, take a good look at all your citations for them. Can you find a more reliable source for key facts? Is what you have a most reliable source, a solid supporting source, or a less reliable source?

Use Color-Coding to Solve Mystery DNA Matches

Last week's article on sorting your DNA matches by grandparent branch was a big hit. Thank you for reading it! Since that DNA exercise was so popular, I've got another one for you. This is especially helpful for those of you whose parents have not DNA tested.

This color-coding exercise comes from a RootsTech presentation by Diahan Southard. It's called, "Shared Matches—The only DNA tool you will ever need."

Choose a Goal No Further Back than a 3rd Great Grandparent

My goal for this exercise is to find descendants of my unknown 3rd great grandmother. All I know is she married Antonio Muollo in the tiny hamlet of Pastene in Italy. To do this, I'll choose the closest match I know who descends from each of my 4 pairs of 1st great grandparents. These should be people who you've already fit into your family tree. Southard calls these cousins your Best Known Matches, or BKMs.

For each BKM, click to view them and find the list of your shared matches. (The location of this list depends on your DNA website.) These are people who match both you and the BKM. Create a color-code on your DNA website for each of your 1st great grandparent pairs. Give the right color to all the shared matches of each BKM. Give it to your BKM from that branch, too.

Divide your DNA matches with this color-coding system and watch all the pieces fall into place.

You can go a step further and give that color to the shared matches of everyone in your BKM's shared match list.

Without using my parents or reusing any BKMs, here are my closest tested cousins for each branch. They're strong matches with longest segments of shared DNA between 35 and 86 cMs.

BKMs by 1st great grandparents branch:

1. Iamarino/Caruso—June 1C1R, 441 cM match, longest segment 51 cM
2. Iamarino/Pilla—Jessica 3C1R, 168 cM match, longest segment 65 cM
3. Leone/Iammucci—Nicky 1C, 793 cM match, longest segment 86 cM
4. Sarracino/Saviano—Christine 2C, 308 cM match, longest segment 35 cM

If that list is confusing, it's because my father's parents were both named Iamarino. They were 3rd cousins.

Use Color-Coding to Filter Your Matches

Next, I'll go back to my main match page and filter my list by only my Sarracino/Saviano branch. The goal is to separate them by my 2nd great grandparent couples. I'll use another color to divide the Sarracino-only matches into a 2nd great grandparent group called Sarracino/Muollo. And I'll divide the Saviano-only matches into a 2nd great grandparent group called Saviano/Consolazio. It's the Sarracino/Muollo group that can help me break through my Muollo brick wall.

Here are the BKMs I chose for my 2nd great grandparent branches:

1. Sarracino/Muollo—Mary, 191 cM match, longest segment 51 cM
2. Saviano/Consolazio—Teresa, 2C1R, 157 cM match, longest segment 35 cM
3. Iamarino/Zeolla—another Teresa, 125 cM match, longest segment 65 cM
4. Pilla/Liguori—Nicole, 59 cM match, longest segment 25 cM
5. Iamarino/diPaola—Keith, 2C, 232 cM match, longest segment 51 cM
6. Caruso/Girardi—Daniel, 2C, 229 cM match, longest segment 42 cM
7. Leone/Pisciotti—Anne, 3C, 97 cM match, longest segment 71 cM
8. Iammucci/Bozza—none! There's only my mother and 1C Nicky.

You May Find Some Surprises

Diahan Southard said this exercise can point out the endogamy in your family tree. Endogamy is when generations of families keep intermarrying. That leaves you with multiple relationships to people. And it's very common in the small, somewhat isolated hill towns that were home to all my ancestors.

I see endogamy right away because my only 1C who's tested, my maternal 1st cousin, is somehow a match to my father. He's a very distant relation to my Dad. They share only 11 cMs in one segment. They each have ancestors from bordering towns, so there are lots of ways they may connect. But I'll probably never find the answer.

If you don't have endogamy, this process should work extremely well for you. You'll wind up with a long list of matches who must belong to a particular branch of your family tree. They've probably been mystery matches because they have no tree or no familiar names.

As a shortcut, I took Southard's advice and looked at Ancestry's ThruLines^® for Muollo branch descendants. There's only one, aside from my 1C and my mom. She's my 3C1R, and she looks remarkably like me. It's possibly she has some Muollo in her ancestry, but she seems to be only on the Sarracino side. If there is a Muollo descendant in my matches, they don't have a public family tree. Still, I will circle back and examine my shared matches with my lookalike 3C1R.

With the Muollo branch at an impasse, I turned my attention to my Consolazio ancestors. My 2nd great grandmother was a Consolazio from the town of Santa Paolina. Many people from that town intermarried with people in the neighboring town of Tufo. And my Tufo vital record collection needs a lot more research.

I chose as my BKM Lynda, my 3C3R, 30 cM match, with a longest segment of 19 cM. I know exactly where she belongs on the Consolazio branch. I marked her and our shared matches with the color for my Consolazio/Zullo branch.

Harvest Your Mystery Matches

Looking at these matches, I moved onto the next step, which is Southard's mantra, "Do genealogy!" I'll choose a shared match with a decent family tree and expand it until it reaches a branch of my tree.

I was digging into one match with a Consolazio family in his tree, but they were from the wrong town. I researched them anyway, and they did, in fact, have roots in my Consolazio town of Santa Paolina. That's a good lesson right there. Don't count someone out because their closer ancestors come from a place you don't know. Dig deeper!

Expert Diahan Southard offers this incredible method for finding exactly the DNA matches you need. I can't praise this method highly enough.

When I did the genealogy, this match whose close family was from another town turned out to be my 3C2R. Our longest shared DNA segment measured 20 cM, which is now very logical.

Our common ancestors are my 4th great grandparents, Gaetano Consolazio and Colomba Ricciardelli. Our shared roots go very deep in Santa Paolina. I color-coded all our shared matches as Consolazio/Zullo.

In the past I marked one of our shared matches as "Consolazio NOT from Santa Paolina." Now I knew I needed to dig into her family tree to bring her roots back to my town. And I did it! I found our connection, which was a mystery before today. She's my 4C1R, and we share the same ancestors—Gaetano Consolazio and Colomba Ricciardelli.

I'd marked another shared match as "lots of Santa Paolina names." Today's exercise brought him to my attention again. Almost instantly, I saw a way into his tree. His grandfather is already in my family tree, but there isn't a cousin connection. When I researched his grandmother, I found our connection. Now this mystery match is my 5C1R.

This is really working! Diahan Southard hit the nail on the head in her presentation. She said, "Stop looking at shared last names, and start looking at shared matches. Then, do genealogy."

This method can solve a ton of your mystery DNA matches. Give it a try!

And speaking of DNA tools:

• This DNA Reference Site Solved a Mystery
• The Leeds Method May Have Solved a Big Family Puzzle
• Combine DNA Tools to Calculate Unknown Relationships

This Spreadsheet Sorts DNA Matches By Branch

If it's at all possible, get your close relatives to take a DNA test that you will manage. Choose someone from your maternal and paternal sides. If you can't test your parents, how about a half-sibling or a 1st cousin? Once they've tested, get ready to see all your DNA matches in a new light.

I'm so glad my parents agreed to take a DNA test. At first I asked them to test so I could more easily see which one of them was related to my DNA matches. Now AncestryDNA is doing that for us with their SideView™ feature. I figured I didn't need this feature. But I was wrong.

While watching videos from RootsTech 2023, one big thing dawned on me. You see, my parents are on each other's match lists. After so much digging, I'm sure they don't have a common ancestor who lived before the late 1700s. But I never thought to look at my parents' SideViews.

Looking at SideView from My Parents' Perspective

The result was pretty exciting. When I looked at Dad in Mom's match list, I saw:

• their relationship falls in the 4th–6th cousin range
• they share 37 cMs
• their longest segment of shared DNA is 14 cMs
• Dad is on Mom's paternal side of the family

That last item is great news because Mom's maternal side of the family dead ends too early. That ancestral hometown was a Papal State in Italy, so it didn't keep civil records before 1861.

When I looked at Mom in Dad's DNA match list, they still shared 37 cMs with the longest segment measuring 14 cMs. But, unfortunately, SideView shows Mom as Unassigned. Ancestry says this could mean there isn't enough information to assign Mom to one of Dad's parents. Your Unassigned matches may get updated in the future, so keep checking.

Analyzing DNA Matches in a New Light

It was Kelli Bergheimer's RootsTech presentation "DNA Misconceptions" that inspired today's project. I created a spreadsheet to compare my DNA matches' relationship to each of my parents and me.

I started by adding the names of my DNA matches in column A of a spreadsheet. I went from the top of the list down to my 44 cM matches. After 44 cMs my matches fell into the Distant Family category. I noted:

• how many cMs we shared
• whether a match is on my maternal side, paternal side, both sides, or unassigned

Then I went through each of my parents' match lists. I captured their matches down to a 44 cM match. If their match wasn't already in my list, I added a new line to the spreadsheet. I checked each of our lists for everyone in column A, noting shared cMs and which side of the family.

When all your maternal DNA matches are on Mom's list, and all your paternals are on Dad's list, you know you've found a cousin.

Next I used colors to show where Mom and I shared a match, where Dad and I shared a match, and where all 3 of us shared a match. In those cases, I added the length of our longest segment of shared DNA with the match in cMs. (If you use AncestryDNA, click your match's number of shared cMs to see what is the longest length.)

I found 10 matches that only Dad has, and 7 matches that only Mom has. All 17 had a strong number of shared cMs. It's possible I share an amount of DNA with them that's too small to make my list. There were no matches in my column that didn't match either Mom or Dad. That qualifies them as true matches and not false positives.

Are Your Matches Really in Your Family Tree?

Kelli Bergheimer speaks about IBD (Identical By Descent) vs. IBS (Identical By State) matches. IBS is also called IBC (Identical By Chance). An IBD match is definitely family—a true cousin. This is someone who fits in your family tree. An IBS/IBC match may or may not be a true cousin. They may share DNA with you because their ancestors' roots are planted so close to yours. And that's one way this spreadsheet helps you understand your matches.

If my parent(s) and I share a match with a good number of cMs, I can be confident that person is a true cousin. I did a bit of research on this concept, and the following numbers vary a bit from place to place. But here's the basic idea.

If the *longest segment* you share with a DNA match is:

• 12 cM or more, you're almost surely true cousins
• about 10 to 12 cM, there's a 90% chance you're true cousins
• between 8 and 10 cM, there's a 50-50 chance you're true cousins
• between 6 and 8 cM, there's less than a 50% chance you're true cousins
• smaller than 6 cM, there's about a 10% chance you're true cousins

Even if you are true cousins, your most recent common ancestor (MRCA) may be too far up your tree to identify. Your odds of finding that MRCA get better and better when you and your match share a segment of 30 cMs or more. If you share a segment of about 16 to 30 cMs, you've got about a 50% chance of finding that MRCA. Any lower and the odds are slim to none. (My parents' longest segment is 14 cMs.)

The message is clear. Don't beat yourself up if you can't identify a match who shares a longest segment of less than 16 cMs with you.

Practical Ways to Use This Spreadsheet

I love this spreadsheet because it can tell me which of my 4 grandparents' lines connect me to a DNA match. I added a column next to the matches' names showing which grandparent(s) is the connection:

• Adamo for Mom's paternal matches
• Mary for Mom's maternal matches
• Pietro for Dad's paternal matches
• Lucy for Dad's maternal matches
• various combinations to cover double matches and unassigned matches

Use your parents' AncestryDNA SideView™ to assign a match to one or more grandparents' branches of the family tree.

This creates a sort of cluster when you sort your spreadsheet by the grandparent column. In my case, if Grandma Mary is their only connection to me, and they aren't a very strong match, I may not find our MRCA. (Again, that's because Mary's line dead ends too early.) If Grandpa Adamo is their only connection, I know they have roots in his hometown of Baselice, Italy. And sometimes a match's last name reassures me that my spreadsheet is correct.

What About Your Closest Tested Relatives?

Getting back to my parents matching each other—I went to DNA Painter to use the Shared cM Project 4.0 tool. I entered the total number of cMs my parents share: 37. The tool says that pedigree collapse or endogamy can affect the results. There's a good chance I have one or both of those things in my family tree. All my ancestors came from a small string of neighboring Italian towns.

Putting pedigree collapse or endogamy aside, my parents have:

• a 50% chance of being 5th, 6th, or 7th cousins, along with several other relationships in that range
• a 19% chance of being half 3rd cousins or 3rd cousins once removed
• an 18% chance of being 4th cousins or half 3rd cousins once removed
• a 10% chance of being half 2nd cousins or 2nd cousins once removed

Based on my extensive family tree, I'm confident my parents are 5th, 6th, or 7th cousins. All I can do hope that their towns' church records go public some day so I can dig into the early 1700s.

If you have very close relatives who've DNA tested, try this type of DNA analysis for yourself. I'm excited to be able to figure out my lower matches who have a stronger connection to one of my parents than to me.

And speaking of DNA matches:

• How to Crunch Your DNA Numbers
• Why Care About Your DNA Matches?
• How to Use the Power of Shared DNA Matches