Mr. Genealogy logo, depicting a shield, typical of those used in family tree coats of arms.
Mr. Genealogy logo, depicting a shield, typical of those used in family tree coats of arms.
Mr. Genealogy logo, depicting a shield, typical of those used in family tree coats of arms.
Contact Me
Menu ☰

 How Accurate Are Online Family Trees?

Can You Trust The Information You Find In Online Family Trees?

This is part of my range of genealogy research information sheets.

Information Sheet 2—How Accurate Are Online Family Trees?

Many people start their family history journey by signing up to one of the main genealogy websites. These are excellent resources, but there is one thing you must be aware of: the various family trees that these contain have been submitted by other users of the sites—they should not be considered as correct. The quality can vary between excellent and complete fiction. It is very easy to accept the information these trees contain into your own, but what are the pitfalls of doing that?

Most user contributed content takes two different forms: there is content where individual users post their own trees versus trees that take a "wiki" style, i.e. they are a single tree where multiple users add their conclusions. It might be argued that a wiki style tree is more accurate, but that isn’t always the case. Any user contributed content is only as good as the researcher who undertook the original work.

Many online platforms allow facts in user trees to be linked to evidence. For example, a burial fact might be linked to a page from a burial register, or at least give an indication on where that fact came from. Academics might refer to this as a citation.

User Contributed Content In Online Family Trees

The best way is to start from the assumption that the tree you have found is wrong, and only accept the information as you have proof. This may sound harsh, but accepting wrong information about your ancestry in recent generations would potentially make anything before that point wrong too, i.e. you could be following the wrong tree. If the wrong great grandfather is followed, then his parents, grandparents and great grandparents, will be the wrong people too—you will be climbing up someone else's family tree.

The only way to know for sure is to look at the information in detail, to examine the evidence, and to see if you believe that this is adequate proof.

What Mistakes Do People Make?

What mistakes do people tend to make when researching their family tree online? By understanding these mistakes, we can also appreciate how incorrect information finds its way into online family trees.

  1. Merged People: The biggest mistake that users make is merging two, or more, people with the same name together. Sometimes it is obvious with a common name that there were many people who shared the name, but even unusual surnames might occur frequently in a small area of the country, and as you go back in time, although the total population was smaller, the pool of given names in use was smaller too. Some of the other errors are a variation on this basic mistake.
  2. Copying Others Without Scrutiny: This is probably the second biggest cause of errors. The online providers make it easy to incorporate information from other users into your own tree. Whilst this is convenient, it also copies things that might be correct, along with things that aren't. It multiples a mistake made by one researcher in the copies. Even if the original researcher notices their mistake and corrects it, the derived trees often don't. Anecdotally, I have noticed a trend here: if many trees contain a certain piece of information, but a few trees contain something else, it is often the minority of trees that have the correct interpretation. It is also these minority of users who tend to have smaller trees. My instinct is because those who are more diligent and take time to research properly, grow their trees slower. Whereas those who copy, tend to build their trees faster (but inaccurately), and more questionable information floats to the top.
  3. Merged Wives: In the past, because of higher rates of mortality at younger ages, and particularly with the risks in pregnancy, it was more common for people to remarry after the death of a spouse. It is normally obvious if a wife remarried, as she will have a different surname, but might not be so obvious if a man's first wife died and then he remarried a lady of the same first name wife. With fewer given names in circulation, this happened more often than you might thing.
  4. Following A Step-Parent: Often the terms "son" and "daughter" are used on a census, even when the child was a step-son or step-daughter. If you rely on census records alone, there might have been time after the birth of a child for the first parent to die and the other parent to remarry. This means that on the next census, the child is a step-child of one of the parents. Not getting the correct records, as many researchers who have shared their trees online haven't, risks unknowingly following a step-parent rather than the biological one.
  5. Grasping At Straws: There comes a point that no family can be followed back further, as the record trail goes cold. Sometimes it is possible for someone with advanced knowledge to overcome this limit, but even they will can only go back so far. That is the point to stop. Many amateurs however find anyone with the right name and continue back. This is generally apparent by the locations on a tree jumping around the country a lot. This phenomenon has a tendency to latch on to a wealthier family with the same surname, rather than the true ancestors. This is because (as a general rule), more earlier records survive for families with more status. Sometimes this is cause by the person who shared their tree only looking at records online (which are far from all sources), and simply referring to records only available offline would have allowed the correct family to be traced.
  6. Failure To Thrive: It is a sad fact that infant mortality was much higher in the past, and many babies died shortly after birth, or before they became an adult. I have seen many occasions where online trees have been created that have found a baptism of someone with the correct name, and assumed it to be their ancestor, but not realised that the poor baby died a few days or weeks later, so could not be the one who went on to be their forebear.
  7. Index Phantom: A researcher who has only built their tree online is more likely to rely on indexed sources, rather than referring back to the original documents (or at least images of them). Indexes are prone to errors. Sometimes those errors are so bad, that someone's name has been indexed incorrectly, creating the appearance that a person existed that never did. I have seen trees where these people who never were have been claimed as ancestors!

Case Study 1—The Father of William Bartholomew Fenton

William Bartholomew Fenton, married Matilda Goddard at St. Mary Newington (then in Surrey), on 7 August 1848. We shall consider this the proven piece of information. The marriage register entry named his father as Thomas Fenton, a printer. William Bartholomew Fenton was baptised at St. Matthew Bethnal Green, on 3 April 1825, with the baptism register stating that his parents were Thomas and Elizabeth Mary Fenton, of Somerford Street, and that Thomas was a printer.

At the time of investigating (January 2023), William Bartholomew Fenton appears in at least 59 public member trees; of these, the majority (35), including a wiki-style tree, have his father as Thomas McLean Fenton (1796-1842), which is wrong. Some have no father, and a few have the father’s name as just Thomas Fenton, with no further information.

Thomas McLean Fenton died in Hastings, in 1842. The death certificate recorded that he was a "gentleman", and he died of consumption. One of the trees even has a copy of this death attached. The trouble is that the status of a gentleman doesn't fit with what was known about the family, who were of more humble folk and lived in the area around London, not 55 miles away on the south coast.

William Bartholomew Fenton's true father (just Thomas Fenton), had died of typhus fever in 1839, at Christ Church Workhouse, Whitechapel. His death certificate confirmed that he was a printer.

For those in any doubt, other records (in the form of poor law documents), make the proof absolute, but all the signs were there to start with: William Bartholomew Fenton's father was only ever recorded as Thomas Fenton (not Thomas McLean Fenton), he was of working class (not a gentleman) and lived in London (not 55 miles away in Hastings). Yet, most trees online have the wrong information.

Case Study 2—The Father of William Abbott (c. 1847-1905)

William Abbott junior, was born on 1 December 1875, in Felling, County Durham. His birth recorded that he was the son of William Abbott, a cooper and Bridget Abbott formerly Gallagher. It is William Abbot (husband of Bridget Gallagher), who is the subject of this case study.

The Abbott family were located together on the 1881 census, in Heworth. This recorded that William Abbott senior, was born in Ireland.

At the time of writing (2023), William and Bridget appear in the region of 100 user contributed trees. Those that have a father for William (around half), have the said father as William Abbott (1818-1869), a chemical labourer. None of these trees have a copy of William and Bridget's marriage.

Obtaining a copy of the marriage certificate clearly stated that William Abbott's father was Patrick Abbott, who had been a cooper, but was by then (1865), deceased.

So why then do so many user contributed trees have the wrong information? Beyond the fact that probably most of them have just copied the information, there are two reasons. Firstly, and most importantly, none of the users have obtained the best document, i.e. the marriage certificate, which gave William’s father’s name and occupation. Secondly, there were two William Abbotts, of a vary similar age, in Heworth, County Durham. Firstly is the "other" William Abbott, found with both his parents on the 1861 and 1871 census in Heworth. Whereas, the William Abbott who married Bridget Gallagher can’t be found in 1861—perhaps he was still in Ireland at that point—and by 1871, was married and enumerated on the census with his wife, not unmarried and living with his widowed mother as the users with the wrong information had!

Summary

In summary, don't accept facts from online trees into your own tree without scrutiny. Apply logical tests based on the evidence, but remember: a fact that appears in many trees doesn't make it more reliable, it just means that it has been copied more.

When looking at a family tree, a professional genealogist can often "smell" when something looks wrong—deliberate mixed metaphors—but they will always obtain the proof needed, so that an accurate family tree is developed.

References