Be more restrictive about the use of a "before editing high quality information" alert.
The story of the Little Boy Who Cried Wolf not only teaches us not to lie, but it also should teach us to be careful about sounding alarms. That said, I think we need to reconsider when the "before editing high-quality information" warning should appear. For example. I am working with William Edward Midgley GNG7-4K1. His Birth field currently says "1877" "Keighley, Yorkshire, England, United Kingdom." The tagged sources show that the year of birth appears on the 1881, 1891, 1901, and 1911 census records as 1877.
The 1939 census shows me that he was born 4 February 1877. Therefore, I would like to edit his birth to be 4 February 1877. When I go to edit "1877", I get a warning about "before editing high-quality information…" I do appreciate the warning, and I agree that before editing ANY information, our change should be supported by a source (this is true all the time). I do understand that a birth of 1877 is supported by four censuses. However, I think that the high quality warning should be reserved for truly high quality information such as a full day, month, year.
My worry is that I have already seen that warning so many times on information that I fully intend to change because I have a full date instead of just a year based on a census (which census is commonly off by a year in either direction), that I already have stopped "seeing" the warning. I already am ignoring its existence and not bothering to read the alert because I've already figured out that it isn't protecting seriously accurate information like a day, month, year, it is just protecting info that matches with rounded off dates on sources.
We already have the quality score at the top of the page that tells me the quality is HIGH. So I already know the current information matches with the sources. Could we reserve the extra "before editing high-quality information" warning for full out dates and locations? If the best we have is a year or just a country, we certainly aren't as high quality as a day/month/year and city. Let's save the high quality warning for full day/month/years so that when a patron sees that alert, the alert gets taken very seriously.
Also, I work with beginners. I fully expect that a beginner would hesitate to add a full day/month/year because there was a "before editing high-quality information" warning.
I like that the high quality score exists the show the sources match the vitals. I wish the "before editing" warning were reserved for the very, very best data.
コメント
-
Totally agree with this request. It is definitely a balancing act trying to get the right phrasing, and the right trigger criteria, but treating a date of just a year from a census as "high quality" instinctively makes me wince. Apart from anything, UK censuses on public release (the 1939 Register excepted) don't actually include a birth-year - that's a calculated value that can be one year out depending on whether the subject has had their birthday in that year or not. And that's disregarding the way some people float their age around.
So yes, I appreciate the balancing act but agree with the OP that this one should be dialled back a bit in these circumstances.
2 -
Thank you everyone for your input, and your patience while we get the algorithms dialed in for this Beta project.
1 -
If I could share a different perspective…several times on some of my well-researched lines, users who were clearly inexperienced travelled through the lines and introduced multiple errors. I'm sure their intentions were good. I got the sense they just didn't know what else to do, so they re-worked lines they shouldn't have touched.
I see these warnings as being mostly geared toward these inexperienced users. I think they will pay more attention than experts like the OP :) I'm sure the algorithms will contine to be refined, but I think I'd prefer to have the warning even in situations where it might not be needed than to skip it when it is needed.
3 -
perhaps warnings are more germane when baptism or birth records are available it would also be helpful to consider DNA evidence in some way in order to consider a source to be of the highest quality.
0 -
@Nancy L Sabo suggested…
"… it would also be helpful to consider DNA evidence in some way in order to consider a source to be of the highest quality … "
I don't know if anyone has brought up DNA in the context of Profile Quality scores but I if they have, I bet there's a whole list of issues. I'd be reluctant to include presence of DNA (DNA matching? Not even sure what one can and / or should look for…) as a measure of quality, not least because it's just a matter of pot luck if anyone on a particular line has been tested, and one shouldn't knock a point off a score if no-one has.
0 -
"I see these warnings as being mostly geared toward these inexperienced users. I think they will pay more attention than experts like the OP :) I'm sure the algorithms will contine to be refined, but I think I'd prefer to have the warning even in situations where it might not be needed than to skip it when it is needed."
EXACTLY!
One of the bigger problems with family trees has been where beginning researchers want to contribute, but they haven't learned to look at evidence /before/ doing an edit. I have one line where there are many Abraham/Abram/Abe males, all living in a general area, all with variations on the same last name, and all descended from the same couple in Maryland in the 1700s. One bad merge, or severing children and spouse can cause havoc. To me, the warning has merit in possibly halting incorrect edits by giving the profile's rating as validation.
1 -
@monnettohio pointed out that
"… One of the bigger problems with family trees has been where beginning researchers want to contribute, but they haven't learned to look at evidence /before/ doing an edit. … "
While I agree whole-heartedly with that statement, shouldn't the solution to that genuine problem start with training beginning-researchers before they mess up the Abraham / Abram / etc lines?
Once upon a time I placed my faith in the User Interface until I had enough bitter experience to persuade me that the other guys were right when they advocated training first.
2 -
"While I agree whole-heartedly with that statement, shouldn't the solution to that genuine problem start with training beginning-researchers they mess up the Abraham / Abram / etc lines?"
In a perfect world, yes. But we don't live in a perfect world.
There needs to be a way to tell the casual user that the profile meets an established fact pattern The wording is a way to do that. I for one like it, and think its way over due.
1
