Data Quality Flags in Data View Popup

Gordon Collett

Interesting new feature. I'm still trying to decide what I think of it, but since it is here it would be really handy to be able to dismiss the flag right here rather than needing to close this popup, click on the Quality Score details link, find the flag there, then dismiss it.

On a second issue, should this be flagged at all? A lot of sources will not have a person's full first and middle name. If a source just has the first name and the profile has the first and middle name is that really a quality issue? Or should the routine recognize that the name in the source matches the first part of the first name in the profile?

My waffling with showing the quality flag here is being torn between the insistence that we take care of this flag right away and the fear that inexperienced user will think the only way to get rid of it is to delete Johnson from Severin's name vs. liking that it does show what flags are influencing each data item.

I do have a question why in this case the name is shown as high quality with the flag being present and why in the following case the name is also shown as high quality even with so many flags:

(I did just dismiss all five flags because none of them were actually problems. They were just differences in recording styles and spelling.)

Gordon Collett

The more of these I find, the more annoying it is getting to not be able to dismiss the flag right then because it is not a problem at all or is an obvious problem in the source rather than the profile:

Please find a way to let us dismiss this right here which would decrease the interruption in work flow the notices cause.

Mary Anna Ebert

@Gordon Collett Thanks for the feedback!

If you put the spelling error or name variation as an alternate name, the warning goes away. This also would indicate to me that someone has reviewed the name. What would you expect as a user? A spelling variation list instead of adding an alternate name?

We are going to look at adding the dismiss function to the conclusion dialog.

Thanks again!

Gordon Collett

I think you would agree that putting in:

as a mis-spelled alternate name would be silly. It clearly has no basis in the original record. Likewise, adding in as alternate names erroneous transcriptions in an index will just create unnecessary clutter and rather worthless junk in the alternate name list. I do add in real alternate names as appropriate and have seen that those are checked by the Data Quality routine before flagging a name conflict and that works quite well.

I think that the best indication that someone has reviewed a name is the reason statement entered when the data quality flag is dismissed. That puts the reason on that Dismissed Research Helps tab along with all the other similar explanations such as for dismissed hints and dismissed possible duplicates.

I think the Alternate Names work just fine the way it is and adding a list of mis-spellings would just be a confusing distraction.

What I would like to see is:

rather than needing to click on a flag to open a second pop up on top of this pop up just to find the Dismiss button. Having the dismiss button right there would also defuse the irritation some users are having now with the current side panel who do not realize that you can click on the flags to dismiss them and so think there is no way to get rid of them.

Would this increase the risk that users would just click through the list and dismiss everything without considering anything? Possibly. More than the current system of dismissing them? Probably not by much.

Gordon Collett

Here is another example of why it would be really nice to just be able to dismiss flags right in the data view popup:

She does have an alternate name of Johanne Brynjulfsdatter. The mangled transcription of Bregjeltors should not be put as an alternate name and should not be put in as a mis-spelling because it is neither.

The actual source shows that the name is clearly Brynjulfsdt.

I'll have to admit that since the original record is so easy to get to since the indexed source is attached and has an image, I have to rank correcting the index at a very low priority. Maybe in a few decades when I run out of everything else that needs doing in Family Tree I'll go back to these. But for now just dismissing the flag as an indexing error is all I have time for.

I do occasionally correct indexes when the errors are particularly egregious and confusing such as the index for one parish register where the two witnesses for the marriages were indexed as the fathers of the brides and grooms even though the column header clearly said "Witnesses" and the register did not contain any parents names at all.

Rhonda Budvarson

@Gordon Collett thank you as always for your really good trouble shooting. A ticket has been created to work on this issue :)