Just the facts

[507]

I believe there are more inaccurate family trees floating around out there on the internet than accurate family trees. I have been tested at 37 markers, My predicted haplogroup is I2b1, and i have very few concrete facts to start a family tree on my paternal side. I'm fairly certain of my father and grandfather but not much past that. What is the best way to get started ? Not using speculation and guess work. Just facts. How much can DNA testing really do ?

[1_mke]

Have you signed up with any of the genealogy sites? Census data can provide fairly easily take you back to the mid 19th century if you have unusual surnames or know where your family is from. This is true in the case of the US, Caneda and Great Britain. Probably other countries as well but those are the ones I know about.

[thetick]

Well it depends what you mean by speculation and guess work. Census records are loaded with errors like typos on dates, stupid census takers calling Ukrainians Russian, just down right lying to cover an underage pregnancy...the list goes on and on.

Most Census and vital records are accurate but there is always some speculation and guess work with them.

[Jomid59]

How much can DNA testing really do ?

[1_mke]

Agreed thetick. Other public records are far from perfect as well so you have to develope a good fuzz filter in order to read/interpret them.

As for how useful DNA testing is, I'd say very with the caveat that short of a fairly close match you'll still need to do a lot of work to figure out the connection.

Small matches (<.12% or so) can be many hundreds of years in the past. My Mom matches exactly the same small segment with three generations of the same family for instance. I don't have that segment at all.

DNA testing is only part of a larger arsenal of tools you need to use to find out the truth. Its not a magic bullet at this point. Give it time though and it should get better as more data will be in the databases. I'm going to be very interested to see how ancestry.com's efforts evolve for instance.

Note, I'm talking primarily about autosomal testing here.

[lgmayka]

I have been tested at 37 markers, My predicted haplogroup is I2b1, and i have very few concrete facts to start a family tree on my paternal side.

Does your Y-DNA Matches page list any matches at 37 markers?

[cokie]

Yes there are many inaccurate family trees on the internet. People just copy them without research and the inaccuracies keep circulating.
When you begin, first ask everyone in your family for information and keep a record in a familytree software program - you can download one from the LDS website FamilySearch for free that will do -

https://www.familysearch.org/

Census are really good but you must crossreference over several years to get a good idea of ages and birthdates etc. since the information isn't always accurate. The WorldGenWeb has resources for virtually every country in the world. Birth, marriage and death records are also valuable. Wills will name family members if you can find them. Historical records such as local history books can provide information if your ancestors were in them.
As far as dna, it's pretty difficult as yet to establish one's ancestry exactly from dna. I have over 100 matches on FTDNA and only two matched surnames.

[1_mke]

In regards to innacuracies, some things I look for...

Is the youngest childs birth date realistic? IE, how old was the mother? If she was eight years old at the time you can be fairly sure there is a problem.

Along the same lines, woment seldom if ever had children when they were in their fifties so look at the ages of the youngest children. I generally drop children that are obviously to old or to young. In some cases there migth have been two or more mothers and that will occasionally be reflected in other versions of a particular genealogy.

People generally didn't move around much, even in colonial times. If you have twelve children look for some sort of logical and consisten progression in terms of their birth places. If one of them was born three states over from where every other one was born than there may be a problem.

Particularly in the case of males, look for death dates that predate by a year or more the birth date of their final listed child. Something is clearly wrong in this case.

Eight year old males do not father children. This is a variation of the mother issue above.

When dealing with census data trust the earlier ages more so than the later ones. I've noticed that both men and women tend to shave a few years off their ages as they get older.

Etc.

Mostly its just common sense and judgement calls. I'm slowly learning that all source material is likely to have some kind of error either by accident or intent. You have to take the body of evidence and make your best guess based on what you have.

[Taz85]

Census records are filled with errors ( spelling of first names and surnames, year of birth, year of immigration). Ive seen errors in all of them. When your going farther back, you are to find errors and there could be alot of guesswork..

[WCoaster]

Mostly its just common sense and judgement calls. I'm slowly learning that all source material is likely to have some kind of error either by accident or intent. You have to take the body of evidence and make your best guess based on what you have.

Agreed! Many trees are inaccurate and/or unsourced, but quite a few provide at least some notes and citations - enough that you can do your own follow-up research.

I like Rootsweb’s WorldConnect trees because they’re so easy to use. The trees are well-organized and free of clutter. There’s an alphabetical index of names you can search, pedigree and descendancy views, etc.

Find the mini-trees on Family Finder helpful, too. They provide a context for the names, and I think most people try to make their trees as accurate as possible.

[sspeters41]

I think we all agree that there are a great number of errors out there. Some are typos, I have made them and I hate it when I do.

What really frustrates me, is when there are errors in fairly recent genealogies and you e-mail the individuals about the errors and give them documention that shows them they have errors and they still will not correct their databases.

When it is brought to my attention that someone disagrees with my information, I will confirm the information and if my information is incorrect, I will correct and thank the person who brought the error to my attention.

I always appreciate corrections and additions to my family trees.

[EdwardRHill]

I like using Ancestry but your right about the tree's. When I first got my results from the Y-DNA test I found out how bad it is. I matched up with Lee's that a different surname than mine. But i had to get started some how and I started looking at tree's with Lee's in it that were in the area of interest.

I found out that 95% of all tree's with Lee's in it had made a connection to the Robert E Lee, Richard Lee line from Virginia, I kid you not. But the trick is to look at tree's that someone has put some effort into. Tree's that have some effort put into them can be a gold mines.

[507]

I match 2 people on 37 of 37 markers and 6 people on 36 of 37 markers. The project coordinator says he is certain without a doubt of the ancestors i am descended from from Pennsylvania in 1682 based only on my dna. Is this possible ? The project has 98 members total. I only know for certain back 2 generations but he says there is no doubt whatsoever about my lineage. How is this possible ?

[EdwardRHill]

I match 2 people on 37 of 37 markers and 6 people on 36 of 37 markers. The project coordinator says he is certain without a doubt of the ancestors i am descended from from Pennsylvania in 1682 based only on my dna. Is this possible ? The project has 98 members total. I only know for certain back 2 generations but he says there is no doubt whatsoever about my lineage. How is this possible ?

Y-DNA is pasted down from father to son over the generation with little to no change. So the test can identify the line your from and give you a idea to which generation you could share a Common Ancestor with your matches.

That along with the information he probably has from your matches makes it possible to say your from that line. You just have to connect the dots. You have points A, B, and he gave you F you just have to fell in the blanks.

[Zaru]

Y-DNA is pasted down from father to son over the generation with little to no change. So the test can identify the line your from and give you a idea to which generation you could share a Common Ancestor with your matches.

That along with the information he probably has from your matches makes it possible to say your from that line. You just have to connect the dots. You have points A, B, and he gave you F you just have to fell in the blanks.

I am not entirely convinced that the mutation rates are correct though. I am PA for a project where no common haplotype has emerged. Although they all look similar with regards to clustering, there is enough of a GD, that according to the accepted explanation, no relationship can exist within a the 400 year range back to the shared immigrant ancestor.

Another thing to keep in mind is that if you have a group like mine, finding the baseline haplotype is a challenge. My group is small enough where I have reached out for trees and have had about a 30% success rate in getting them. Comparing and contrasting common lines is key to finding out where the hiccups might lie. In the case of paternal lineage, the hiccup can only occur with either the mother, or an adoption. The father having affair with another woman will still have his Y-DNA. So look for the clues that are associated with the mother.