Ancestry Public Trees versus Private -an invited Referendum

[peter]

The only problem with that Peter is those with the worst trees are not aware of the inadequacies lurking within them.

Bob, even assuming you're correct, you've forgotten how this discussion began - it was with an article in the newsletter where I pointed out that Ancestry had advised beginners to post trees which they knew contained inaccuracies, and that this advice was repeated in the FamilySearch wiki.

There will always be people who wrongly believe their version is correct, but are you still disagreeing with my suggestion that people who upload a speculative tree should choose an appropriate title for that tree, as a warning to others?

 

[peter]

I think most reasonably experienced researchers will be aware of the possibility of inaccuracies but will nevertheless take care to ensure their tree is as accurate as possible.

Of course - but given that there are currently only 2.7 million Ancestry subscribers, and that not all of them are "reasonably experienced", what proportion of the 90 million trees at Ancestry do you suppose are "as accurate as possible"?

If all researchers (not only genealogists) kept their research private to avoid any possibility of publishing anything inaccurate, not only would we all miss out on the benefits of research, but researchers themselves will also miss out on valuable peer review.

Nobody has suggested that all research should be kept private. But even if it were, so long as it was searchable there would be nothing to prevent cousins connecting and exchanging information - who better to peer review the evidence than cousins who have begun there researches from a different viewpoint?

Surely the real problem is that the platform that Ancestry provide is not fit for purpose? Peer review sounds wonderful, until you consider how ineffective it can be. No matter how inaccurate a tree is, the owner can ignore comments and suggestions, leaving the tree online for other less wary users to trip over. And that's even assuming the tree owner is still alive and that Ancestry have a valid email address for them.

Consider this: eBay wasn't the first online auction site, but they succeeded where others failed because they had a system of peer review which worked, and where the ratings were highly visible. All you can do at Ancestry, other than write to the tree owner, is post a comment against an individual on a tree - and that won't show up in search results.

 

[Bob Spiers]

There will always be people who wrongly believe their version is correct, but are you still disagreeing with my suggestion that people who upload a speculative tree should choose an appropriate title for that tree, as a warning to others?

Not really although not one so flippant as the example you used in your Newsletter, and I dislike the term 'warning'. There is nothing wrong with the well established term 'BETA' which all should understand means new and not fully tested. I also use 'TENTATIVE a lot, especially on pages in my Tribal pages to let family know facts are still being researched, and sometimes attach the word to an Ancestry Tree title.

For example I set up a new public tree in Ancestry last year - something I promised to do when my grandson's Bermudan in-laws expressed an interest in exploring their own British/Portuguese/American roots when on a visit to the UK (accompanied by my grandson, his wife and baby -their grandson, my great grandson) My provisional title for the Tree was XXXX-YYYYY (Bermuda) 'tentative'. It remained that way for best part of a year before I was sufficiently happy to remove the tentative label. I am currently in the process of finalising their Tribal pages Tree and they have been providing photos which will be incorporated.

Most of my (Ancestry) Trees are set up by importing gedcoms from previous research. I give these a simple title usually showing the main Grandparent surnames. I do however use the 'Tree Description' label to explain in more detail what the Tree is about and usually at what level researched.

 

[Pauline]

what proportion of the 90 million trees at Ancestry do you suppose are "as accurate as possible"?

I wouldn't like to hazard a guess, but whatever proportion it is I am not sure it is grounds for commending people who opt to make their trees private. As you correctly said before, it is a personal choice, and it is not because of an awareness of possible inaccuracies that I have currently opted to make my tree private.

But even if it were, so long as it was searchable there would be nothing to prevent cousins connecting and exchanging information - who better to peer review the evidence than cousins who have begun there researches from a different viewpoint?

I did acknowledge that having a private tree doesn't preclude peer review.

Surely the real problem is that the platform that Ancestry provide is not fit for purpose?

I wouldn't go as far as to say that, but I do prefer using my own genealogy website for my tree.

 

[Tim]

Wouldn't it be great if you could a thumbs up or a thumbs down against a fact on a person in someone else's tree, and obviously leave a comment as to why.

 

[Tim]

it was with an article in the newsletter where I pointed out that Ancestry had advised beginners to post trees which they knew contained inaccuracies, and that this advice was repeated in the FamilySearch wiki.

I know this is what the article says, but it is slightly out of context. What I think they are trying to tell beginners is that if you add a rough DOB or a range, then this allows the searches to propose some suggestions which should lead to the correct person. And I think this is fine, and to be honest, I think we all do that anyway? e.g. When you get a christening and it's dated 1803 and gives the parents names, when I add the parents I don't leave the DOB blank, I add in a range something like BET 1760 and 1785. This helps when doing searches and surely must be standard practise by us all?

However to your point, if they just guessed dates and left them in their tree and never corrected them then they would pretty quickly run out of people and facts to add.

 

[peter]

There is nothing wrong with the well established term 'BETA' which all should understand means new and not fully tested

That may be the case nowadays but, when I was in the industry, beta software had been fully-tested internally. What you're referring to is what I would have called alpha, not beta.

A LostCousins member who teaches family history advises his students to prefix the names of their trees with the word 'Test'. Personally I would prefer 'Draft', but there could well be an even better term.

What I think they are trying to tell beginners is that if you add a rough DOB or a range, then this allows the searches to propose some suggestions which should lead to the correct person.

Should? Surely it's might? There's no guarantee that the record will be in Ancestry's database (maybe Findmypast have it), and even if it is, the user won't necessarily be able to identify the correct record. If the name is a common one there could be dozens of alternatives. A lot will depend on the level of enthusiasm of the beginner , as well as how much time they have (and how much patience they have been blessed with). Remember too that not everyone is as smart as LostCousins members - just look at the number of people who can't even spell Ancestry.

In any case, is this really the way to teach beginners how to do research? Even if they manage to pick the correct record, what have they actually learned? I believe in teaching people techniques, not spoon-feeding them with answers - I frequently spend 20 minutes explaining to a member how to do something rather than simply giving them the answer, which might only take 20 seconds.

 

[Tim]

So very true Peter. As the saying goes, "Give a man a fish, and he can eat a meal. Teach how to fish and he can eat for life."

 

[Bob Spiers]

I know this is what the article says, but it is slightly out of context. What I think they are trying to tell beginners is that if you add a rough DOB or a range, then this allows the searches to propose some suggestions which should lead to the correct person. And I think this is fine, and to be honest, I think we all do that anyway? e.g. When you get a christening and it's dated 1803 and gives the parents names, when I add the parents I don't leave the DOB blank, I add in a range something like BET 1760 and 1785.

How remarkable is that, seconds before receiving an Alert which turned out to be your post Tim, I was about to respond to Peter's mention of 'inaccuracies' and had just got as far as copying part of the script when I stopped to read the new alert. I'm glad I did or our comments would have crossed and believe it or not - give or take - I would have made the same points.

For as long as I have been using Ancestry they have suggested when conducting a Search, if vague on details -names, dates, places, what have you - then make a best guess and see where that produces. Such input is intended to act as a springboard which -along with wild cards- give Ancestry's search algorithms a chance to come up with something. Even though at times the outcome can introduce more vagueness that you had to start with! FMP on the other hand would prefer you enter as little as possible and not complain when it provides over 1000 hits for you to sift through. (OK both remarks are said with tongue in cheek as both search techniques have their uses, sometimes producing excellent results).

I am no biblical scholar but I am aware of the phrase ..."out of the darkness comes light" (or something similar) so I am sure the same can be said for vagueness giving way to clarity; one can hope anyway.

 

[Pauline]

So very true Peter. As the saying goes, "Give a man a fish, and he can eat a meal. Teach how to fish and he can eat for life."

Of course, some people are just looking for a quick bite to eat, and don’t want to learn how to fish.

 

[Bob Spiers]

Wouldn't it be great if you could a thumbs up or a thumbs down against a fact on a person in someone else's tree, and obviously leave a comment as to why.

No not really as it would then be like 'Trip Advisor' and we all know how misleading and biased some of the comments left by fellow travellers can be. Perhaps, whilst still in a biblical mood, 'let he who is without sin' be our watchword.

 

[peter]

Wouldn't it be great if you could a thumbs up or a thumbs down against a fact on a person in someone else's tree, and obviously leave a comment as to why.

No not really as it would then be like 'Trip Advisor' and we all know how misleading and biased some of the comments left by fellow travellers can be. Perhaps, whilst still in a biblical mood, 'let he who is without sin' be our watchword.

That's not a fair comparison - an experience in an hotel or restaurant can vary enormously according to when you go, who you go with, what you ordered, who served you, whether or not you were in a hurry, what your expectations were (and how realistic they were), and so on.

Nevertheless I'd rather be able to comment on a tree than an individual. Unless the comments feed through to the Search results users are still going to waste time looking at duff trees (there may not be any comments against the records they're looking at).

But I'd argue that equally important is whether the tree owner responds to enquiries, and whether they deal appropriately with the feedback they get. After all, someone who does that is going to end up with a pretty accurate tree eventually.

Most important of all, I'd like to know whether someone who turns up in a tree search is actually a cousin of mine - and you currently can't tell from the search results whether or not that's the case. Even when you look at the tree you can't always be sure.

 

[peter]

For as long as I have been using Ancestry they have suggested when conducting a Search, if vague on details -names, dates, places, what have you - then make a best guess and see where that produces. Such input is intended to act as a springboard which -along with wild cards- give Ancestry's search algorithms a chance to come up with something. Even though at times the outcome can introduce more vagueness that you had to start with!

There's a big difference between searching and uploading a tree. Fuzzy matching makes a lot of sense when you're searching because even if you know (or think you know) what the record should say, it could have been recorded and/or transcribed differently. When you put something in your tree that's a guess you need to make it clear that's what you've done, otherwise you could mislead yourself.

Of course there are unknowns and guesses on the fringes of our trees - how can there not be? If my ancestor died in late 1835 and her age was shown as 32 I'll pencil in 1803 as her year of birth, even if I haven't found her baptism. That won't mislead anyone, but if I attached an incorrect source it might. For example, one of my ancestors is missing from the 1841 Census, but several Ancestry users have attached a record from that census which clearly relates to a different family - the father's name is right, but his birthplace is wrong, his wife's name is wrong, and the children's names and ages are all wrong.

OK, in that case it's obviously wrong, so nobody who actually looks at the record critically is going to be misled. But the fact that so many people have the same incorrect census record in their tree shows that a lot of users aren't looking too closely. If they don't spot such a glaring error they're clearly not going to notice less obvious errors.

 

[Bob Spiers]

...nobody who actually looks at the record critically is going to be misled. But the fact that so many people have the same incorrect census record in their tree shows that a lot of users aren't looking too closely. If they don't spot such a glaring error they're clearly not going to notice less obvious errors.

I agree, and time and again when I discover something is screamingly obviously wrong, I still find it amazing when others latch on to the same information. That has to be a red warning light to leave such Trees alone. But there is the occasional oasis in the desert when I spot a Tree that doesn't follow suit... THAT is the Tree* to examine further:

* I would think it far easier if the tree was a Public Tree one, to test your instinct and move on if wrong. It is certainly more convenient than having to send the tree owner a message asking for permission to view, or whatever.

If a more detailed examination shows my instinct was right it can be rewarding to discover a new or improved fact (additional family member, birth places, dates etc) than the one(s) currently in my Tree. (Not taken entirely by trust I might add but something to note and research further).

Even though I admit there are tree owners whose trees are farcical at best, I have learned over time to pass them by and get on with my own research. I just endorse and make contact with tree owners who at least care about the information they show. (Of course if they don't respond then they get demoted and I move on).

 

[peter]

But there is the occasional oasis in the desert when I spot a Tree that doesn't follow suit... THAT is the Tree* to examine further:

Agreed.

* I would think it far easier if the tree was a Public Tree one, to test your instinct and move on if wrong. It is certainly more convenient than having to send the tree owner a message asking for permission to view, or whatever.

I agree, it's much quicker and easier when the other person has a public tree. I am not against public trees in principle, only in practice - because of the poor implementations. Monkeys with typewriters could have designed a better system.

 

[Bob Spiers]

I agree, it's much quicker and easier when the other person has a public tree

You can say that again (figuratively speaking) but as I have only just got round to checking on DNA matches for my wife's DNA results (received January) I can truly relate to the whole sorry mess of... (a) 'No family tree'; (b) '(tree unavailable)'(why?); (c) 'locked (private) trees'(very few actually) and whilst the remainder (d) were 'public trees' how disappointing to find so many with but a handful of names. So first impressions not at all inspiring. Worst of all my wife said "hardly worth the money", but as it was my gift to her I just answered and said it depends on the outcome?

I had only just read the search techniques set out in your last Newsletter (I was not ready for DNA when it was first aired in Aug 17) - very useful by the way - but first I had to answer further questions from my wife. Why were there so many 'No family tree' results? Why are there some with less than 10 names (one with 5)? What do you mean you have to make contact to view a private tree ? (I told her not to go down that road). In the end I said I will need to explore further, and see what comes from checking those I can view

You may understand my disappointment when I give the statistics of the first page of matches:

3rd Cousin = 3 results: 2 x No family tree, one public tree (c1800 people)
4th Cousin = First 25 only: 12 x No family tree -1 (tree unavailable) - 2 x private: 1 x 300 people(starred) and 1 x 3000 people - (remainder public): 3 less than 10 people - 1 less than 20 people (one was starred) - 1 x 68 people - 1 x 127 people - 2 x c 300 people, 1 x 800 people & 1 x 1000 people (the latter starred)

(Initial result): The public tree designated 3rd Cousin with 1800+ names turned out to be completely unfathomable and found nothing using the name/place searches. (Have put that one on ice).

Leaving aside the private trees for the time being I can report the starred tree with only 15 names proved a 3rd cousin match using the name search technique. It found her great x 2 grandparents which happen to my wife's also; hers paternal, the owner's maternal. Amazing 15 names and I found a match! I have just messaged the tree owner on behalf of my wife and wonder what she'll make of my wife's Tree which (although tiny by the side of my own Ancestry Tree) at least boasts 850 names, and would think a quarter of those are likely to be of interest.

The other success came from the public tree with 1000 names (starred you recall) which also latched on to a 3rd cousin relationship; this time a reversal, my wife's maternal line and the tree owner's paternal. A rewarding find because of the greater volume of names and further exploration required. Owner duly contacted.

My next step is to contact the starred private tree with 300 plus names (a man). Hopefully this will also be rewarding providing I get a response. The only other private tree with 3000 names (a lady) will, because of the volume of names, also go onto my contact short list.

Names and place searching -other than the two successes so far reported -turned up very little and none at all for the major names, except I should add 'Smith' in Northamptonshire which, not surprisingly, found a myriad of names, but none favourably so far.

I will deal with page two and onwards later...but almost forgot from my list of 'gripes' the number of tree owners with 'Last logged in' showing dates in 2017. At least the two successful matches had very recent dates and so will concentrate on those who at least last logged in this year.

I supposed this posting is a 'tad' off topic, but as you will see it airs the public versus private topic, so perhaps OK.

 

[peter]

I supposed this posting is a 'tad' off topic, but as you will see it airs the public versus private topic, so perhaps OK.

Staying off topic I shall relate my experience from this weekend. I found two new cousins who had public trees - though neither has replied to my messages so far. One of the trees only had 40 entries, but it was more than enough.

I investigated another DNA match with someone who had a private tree, they came back to me almost immediately and gave me access to their tree. We still don't know how we're related, but that's because despite sharing a substantial chunk of DNA, the common ancestors are on the other side of our 'brick walls'. No instant gratification, but a pot of gold in prospect at the end of the rainbow.

Reading Bob's comments reminded me how important it is to stick to the strategies in the Masterclass, and not get diverted. Focusing on the first page of matches (or indeed, any page) is likely to lead to disappointment.

For those not familiar with Ancestry DNA starred matches are those that the user has flagged - they're not chosen by Ancestry.

 

[PhilGee]

....
4th Cousin = First 25 only: 12 x No family tree -1 (tree unavailable)
...

Usually, when I get "tree unavailable" it clears with a refresh.

...but almost forgot from my list of 'gripes' the number of tree owners with 'Last logged in' showing dates in 2017.
...

What, no 'Last logged in 2014' ?

Phil

 

[peter]

Usually, when I get "tree unavailable" it clears with a refresh.

In my experience refreshing doesn't always work but clicking on the match generally does. I think it's all to do with the way Ancestry update their database 'on the fly'.

 

[Bob Spiers]

For those not familiar with Ancestry DNA starred matches are those that the user has flagged - they're not chosen by Ancestry.

Yes quite true, I'd forgotten I had a look at her results a day or so after they were received. I believe the ones I starred seemed to have the most potential - likely by main name matches and presumably that applied to the private tree also?

Focusing on the first page of matches (or indeed, any page) is likely to lead to disappointment.

One has to start somewhere and focussing on the first page seemed as good as any. I shall get around to the others in due time, or they'll get round to me via my searching.

What, no 'Last logged in 2014' ?

Well not on the first page, but I spotted a 2015 on the second page, and I had fair scattering of 2014's in my own result matches.