Quality score for those who died as infants
Today I created a profile for Ethel Mary Ibbotson GRZR-1ZT, who was born in 1892 and died in 1893. The only records I could find for her are a civil birth registration and a civil death registration. I attached and tagged both.
However, she still only has a medium score because she has no burial information, and no month for her birth or death.
The only way I could get exact birth and death dates would be to purchase the civil registration images, which I'm not planning to do. And maybe some day FamilySearch will film the parish registers containing her baptism and burial (if they exist), so they could potentially be added in the future. But for now, her record is as good as I can make it.
I'd like to hear others' opinions. All things considered, should she have a high score instead of a medium score?
項留言
-
@kathryngz My similar post from a couple of days ago.
1 -
There is one thing that I wish was clearly evident with the quality score. The goal in FamilyTree is to have the most complete and most accurate profile possible, not a profile that is totally complete and therefore inaccurate, and there is absolutely nothing wrong with having a medium or even low quality score.
I can easily picture a profile for a boy who died a couple of days old, before being named, who was only mentioned in someone's journal who did not give any indication of the birth or death dates and nothing can be inferred beyond the year in a situation where there are not and never will be any other records for that child. The record will have a last name, a birth and death year, and one non-indexed source. That is all it would ever have so likely would have a quality score of low even will all issues dismissed. I would likely in this situtation add an Alert Note to the effect that a low score does not necessarily mean that there is anything wrong with the profile and that the profile as it is will never be improved.
The goal in Family Tree should be accuracy, not a high quality sore.
4 -
@Gordon Collett , thanks for sharing your thoughts!
I'm not sure I understand what you meant above when you said "not a profile that is totally complete and therefore inaccurate." Could you elaborate?Your post highlights the challenge of implementing some kind of quality measurement. I think most users would think that there is something wrong with the profile that has a low quality score. As you pointed out, it will be important to communicate exactly what the quality score means.
0 -
"not a profile that is totally complete and therefore inaccurate.": Some people can't seem to bear a blank line. Don't have a christening date? It must have been shortly after birth so just put "about (shortly after birth)." Don't have a marriage date? Put "about (two years before birth of the only known child who might have been their 10th child)." Don't have a death date but was alive in the 1930 census? Just put "After 1930."
Every single field in a profile can be filled in with guesses, calculations, and approximations. But that is not something the Quality Score should encourage. Better a nearly blank profile with only source-established information than a completely filled out profile with only guesses.
5 -
I strongly agree with Gordon's comments here. I do worry that in an effort to get that Quality Score from a LOW or MEDIUM to a HIGH many users will feel a need to complete all the Vitals boxes and add sources that probably relate to the individual in question. Not that some users don't do that regardless (of QS considerations), of course. I see no need to guess birth dates based on an assumption that the bride was 21 and groom 25 when married, for example. Also, without any evidence, the birth year and place fields are completed based again on an assumption: in this case, that it took place in the same year and parish as the christening event. Even the year of death is sometimes entered as one that falls, say, seventy years after the birth, or (as in Gordon's example) an "after" date, which based purely on the date of (census or otherwise) the last record the user has found for the individual.
I hope it will be clearly pointed out that the PQS is merely a guide and there is no element of failure on the part of the user / researcher if they should find it impossible to raise a profile to a HIGH rating, in spite of their efforts. However carefully the developers work to improve the criteria for establishing a "reliable" PQS, there will still be some factors that will cause this to be lower (or higher) than the researcher will believe it should be.
In summary, please don't make guesses, but only input data (and add sources) on the basis of strong evidence that relate to facts, rather than assumptions. Surely that last thing we want to see from this PQS project is that it leads to less reliable details being added to the Family Tree profiles, especially if certain users decide to adopt a strategy for "bumping up" scores.
3 -
Absolutely @Paul W. I dread to think some contributors would try to game the system to increase scores. You may recall there was a thread not long ago about adding games to FamilySearch to get the younger generation interested.
1 -
Thank you everyone for your input. Scoring tends to bring out the challenge to perfect something. Unfortunately this isn't always possible or good. The ratings are meant to give a heads up that something may need a review, or not. Thank you for your patience while this beta product is improved. And again, thank you for your discussion and ideas :)
2 -
This is an interesting thread — my views are quite the opposite. Low scores are to be expected. Very few families are blessed with an abundance of original documentation documenting the lives of their ancestors. My concern is about high scores for profiles that are not well-documented. For example Joseph Hewes LCYM-ZXW which has more than 21 errors — 10 outright false entries — e.g. two different dates of birth both of which are wrong and 11 inconsistent facts — e.g. a date of birth that has a source which gives a different date of birth.
2
