How do we evaluate the published work of other genealogists? That's a very good question, whether we are talking about a book publication, or someone's online tree on services such as Ancestry or RootsWeb or others which allow the publication of genealogies that are owned by one individual.
One place to turn for an answer is the Board for Certification of Genealogists. Their website describes their manual, Genealogy Standards, as follows:
its eighty-three standards contribute to the level of credibility in
genealogy called the Genealogical Proof Standard (GPS).
The five elements of the GPS are:
- a reasonably exhaustive search;
- complete and accurate source citations;
- analysis and correlation of the collected information;
- resolution of any conflicting evidence; and
- a soundly reasoned, coherently written conclusion.
Another metric we can use to judge these works and other evidence is Elizabeth Shown Mills' Evidence Analysis Process Map found in her book Evidence Explained -- you can view it on her website, on this page:
QuickLesson 17: The Evidence Analysis Process Map. (Copy archived on 10 May 2018 by the Internet Archive's Wayback Machine.)
The Process Map has 9 elements arranged in three columns, and sometimes you'll see people refer to it by the nickname "the 3 x 3".
Are the sources:
- Original (e.g. the record at the first time it was written)
- Derivative (e.g. someone's index of town records)
- Authored (a local town history)
Is the information:
- Primary (firsthand knowledge)
- Secondary (secondhand knowledge)
- Undetermined (where we can't tell or it isn't known)
Is the evidence:
- Direct (does it answer the question you have about the person)
- Indirect (information which does not directly answer your question, but can be combined with other information to prove your case)
- Negative (the absence of information which should be there -- not to be confused with negative findings, which might simply be that we didn't search well enough or that something was poorly indexed)
Using these as our yardsticks, can we tell how valid the work might be? Does the author give us a list of sources that were consulted? Is there enough transparency so you can follow along, look at the sources consulted, and evaluate the information for yourself?
Example One: the famous family history
Delmar Rial Lowell wrote a 2-volume genealogy, The Historic Genealogy of the Lowells of America from 1639 to 1899. This authored work was researched by Lowell and his father over twenty-five years, and is mostly un-sourced, although there are some references scattered through out the text. To give you an idea of the scope of the problem -- there are over 2000 individuals listed in the book. A search for "town records" has 14 results. Not all collections of town records will say "town records" in the title, but that gives you a rough idea of how many of the 2000+ people have a note next to their entry, saying which town's records the dates came from.
It is difficult for me to say how accurate this two-volume history is as a whole, when it is the result of 25 years of research, and contains thousands of individuals. Even if we are told that Rev. Lowell and his father had found direct evidence from primary records for every connection in the book, we still don't know if that information was from original records or from derivatives (such as copies of the original records).
The brute-force way to answer the question of how accurate his books are would be to re-do all 25 years of his research myself -- we have no proof statements for any of the individuals. I could get a general idea of how well-regarded the book is by looking for book reviews, or by seeing how many times his book was cited by other genealogists. But any time you count citations, there is an inherent bias towards older published works, simply because they've been around longer and have had more chance to be looked at than more recent books. So you can't just say "more people cite this work so it must be good".
However, it is easier to answer the question you really want to know, which is how accurate the information is for the specific persons / relationships you are trying to establish (a much smaller and easier-to-manage problem than evaluating the accuracy of the entire book).
If you can't answer how the author knows a particular piece of information in their published genealogy, you don't know the quality of the information, so you can't answer how accurate the information might be. If the authors don't cite their sources, we can only use these published works as clues to what original records might exist.
Check your authored genealogies for source lists, bibliographies, footnotes, and other clues that might give an indication of what types of records they looked at. Once you find the original records, derivative works, or other authored works, you can sometimes guess what they might have looked at.
Example Two: the Family Bible
Family Bibles often have records of births, marriages, and deaths. Look at the publication date of the Bible as a clue about when the entries might have been made. If the Bible was published in 1850, then you know that all the entries before 1850 must have been entered at a later date -- a birth which takes place in 1790 could not have been recorded in that Bible in 1790 because that Bible didn't physically exist yet. If all the entries in the Bible were written in the same handwriting, then it's possible that all of the entries date from the time of the most recent event. This seems obvious once it is pointed out to us, but it's easy to get lost in the excitement of finding the information, and skip over the step of examining the evidence.
Let's say you hit the jackpot. You find an article in the NGS Quarterly by a well-respected, Board-Certified Genealogist, with a gorgeous proof statement that answers the very question you want to know. Are you home free?
No, because just like the Bible record I mentioned earlier, that article and proof statement are simply a snapshot of what that genealogist's analysis was at the time the article was written. So don't forget to look for follow-up articles and new evidence that might have been discovered since the article was written.
For a recent case study, see Hillary Clinton Family Tree a Wake-Up Call for Genealogy by Megan Smolenyak Smolenyak, where Megan explains how she found dozens of people had made a wrong identification in part of Hilary's tree.