Has anyone found this has helped their research?
It seems like many in this group have started using this prototype as they go along in their regular day-to-day research.
In what ways has it helped you already?
コメント
-
I'm using it as an extra review. Before I leave a profile, I've started running the Score Summary to see if there is something else I need to do or something I may need to correct before moving to another part of the family or tree.
4 -
As I mentioned in another post, when I randomly tested this on a relative it flagged a residence under Other Information that another user had gotten from the US Public Records database which had a time span for that residence starting ten years after he had died. I hate to admit it, but I can get kind of slip-shod with Other Information added by other people. I don't think I would have noticed this error or gotten it deleted without this routine.
I think this routine has great potential to really help keep profiles in good shape. That is why I am harping of the need to have a very high signal (true problems) to noise (non-problems) ratio so that kernels of truly useful information are not hidden in a ton of meaningless chaff.
4 -
Excellent observation Gordon - I like the high signal versus just noise comparison.
In fact, I have been meaning to post an idea that crossed my mind. Instead of the "5 stars" concept, perhaps a traffic signal - red/amber/green - would be easier to follow/process.
We've talked about not deleting good info just to reach 5 stars. Maybe 5 stars is too high a bar for a new project. Stop/slow/go might be a good option.
4 -
I have been curious how different quirks of information show in the Summary. So I have been looking at my most intricate research to see how everything shows up. The information on the Summary has led me to recheck sources that were or were not attached, confirm data, and check out any unusual statements. All of that is good, however, I question how useful this will be if I always have to go check for this information. I am curious how we would access this Summary. Will it show up as part of the Data Problems? Will it show as a different icon/color from the ones that currently show on the various versions of the Family Tree.
Would it show in the Task List that is available on the mobile app? I would love for the task list to include this Summary, Data Problems and Research Suggestions as well as the Record Hints and Ordinances (green temples). Because otherwise how will I know to check this?
Will this be a scoring that appears on the Person Page? That might put too much emphasis on the 'gotta have 5 stars' attitude that would drive incomplete or incorrect information in order to attain 5 stars....
2 -
I'm sure your developers are too aware of how some people react to new features. However, I'll mention anyway that when this is released you are certain to get comments similar to these about Research Helps: https://www.familysearch.org/tree/person/ordinances/LVQZ-VX5
Particularly regarding the items such as "The burial conclusion is missing a date and place." True and needed to explain why the score is low. But certain users are going to be offended by such an obvious statement. Proper placement on the Person page, full explanation of why statements are present, and the ability for a user to somehow indicate "I've looked at all of this already," will be very important.
2 -
I totally love this idea! I can see great benefits to my personal research.
0 -
To respond to this question, I decided to document a thorough look at what this feature shows. I started with my grandmother and went three generations back from her. In the following spreadsheet the colors mean the following:
Green: Issues resolved with minor editing or tagging sources. This has improved the profiles.
Black: Mainly issues that reflect reality. Some sources like census records have approximate birth years based on approximate ages. Old records rarely say when a funeral was held. Yes, they all died to old. Family information passed down through the generations that everyone knew sometimes does not have sources. Many of these I would just dismiss when we are able to do that. The rest should lead to new research efforts.
Blue: On a Read Only person so there is nothing I can do about them.
Red: Problems that have baffled decades of research of multiple researches that I don't feel competent to even start on. I am doubtful that the sources that are there are for the right person but can't support that thought with anything concrete.
One column is blank because I got an error every time I tried to check her.
There is a lot of green. Many of the black would lead to trying again to improve the situation. So overall there is a definite benefit.
2 -
Wow. Thank you everyone for your insights.
We will work on improving the signal to noise ratio.
I also note that there is no green ink in the Person Data is Conflict Free.
0 -
As far as rechecking information you have already checked, we are looking into making these issues dismissible.
We are trying to figure out how to harmonize this experience with the research suggestions experience.
I'll bring up the untagged conclusions as a possible item for the mobile task list. Any others that seem straight forward enough to put on a task list?
0 -
Right, no green ink in the Person Data is Conflict Free section because all the conflicts are the "very old" flags which are all based on accurate data. Since there is nothing to correct in the profile data, there was nothing to mark green.
1 -
I'll bring up the untagged conclusions as a possible item for the mobile task list. Any others that seem straight forward enough to put on a task list?
Untagged Conclusion is straightforward, but you need to qualify it; are there any sources actually attached to the record? If there are no sources attached, it should show you the Research Help suggestion of "No sources attached." rather than that a conclusion has no tagged sources.
Personally I'd love to see all of these on the mobile task list. Perhaps filterable? Completing data, Checking data, Temple or Hints?
1 -
Update: The engineers hope to make this into a list format after they build a more robust system. Thank you for your patience :)
0 -
I was going through some profile quality hints, and I noticed this. I was all prepared to write an error report saying that the algorithm missed the city, but I decided to look farther into why the algorithm might've come up with that report, and it turns out that the marriage record only lists the marriage county. The family lived in Tacoma, so I hadn't suspected a thing, but it could've been anywhere in the county.
Good job, engineers!
2
