Problematic Project with many mutual (wrong) IDs - what to do?

Here’s the situation:
there is a project going on right now containing more than 1,000 observations already, and many of them are falsely IDed. Some members of the project (more than 20 users in total) ID many of their colleague’s observations, so that they become Research Grade with 5-6 supporting IDs.
With the result that even two disagreements are not enough to revert them.
And even after the (maverick) disagreement was made, other members will leave another vote for the group

I wrote to the two project admins directly, but no responses.

What should be done in such a case?

  • Should I link to the project here and ask for a concerted action from forum members to help clean it up (Hong Kong-style)?
  • should I alert the admins at help@inaturalist?
  • should I flag the project (well, the project itself is fine, it’s just the inappropriate behavior of project members)
  • should I just wait until these issues are resolved in the months (or years) to come?
1 Like

Contacting admins sounds like a good first step to do, asking others to help would also be a partial solution of current state of situation (even if users are suspened, I think their ids stay?).

2 Likes

This is kind of like fighting a battle that you know you will lose. At least for the HK projects, the rate of uploads far exceeds that of the rate in which we respond accordingly.

1 Like

that’s hardly comparable - it is a time restricted project with few members. The issue here is that you will need several IDers per observation if the users turn out to be non-responsive

1 Like

Yes, you guys have it a lot worse.

Quite a few ‘Lucilia sericata’ and ‘Calliphora vicina’.

First I’d ID some of the observations, maybe a lot if time, and leave ID basis comments or simply tag the usernames for the wrong IDs. Some will notice eventually, and notice your IDs. The misidentifications are like a domino effect through a community, but the corrections can be too. I’d only resort to other measures if this didn’t work, after some time. Other people may come by and add correct IDs too. It also wouldn’t make sense to flag anything.

Haven’t at least a good deal of those been corrected by now? I checked the Hong Kong bees and wasps, although maybe those were less affected. I at least don’t know of any cases in other locations either where this occurs but can’t be at least mostly contained by adding correct IDs. Or are there examples where misidentification RG obs. permanently outnumber correct RG obs.?

If this is a school project, from experience, I’d say that they will probably resent you for your IDs, or just ignore them.

3 Likes

I see, I’d still correct them anyway. Also, I do think it’s a fine idea to share with other identifiers or recruit help, as the poster did. I just would only notify staff it it was very large and uncontainable, although I don’t know how big this one is. I wouldn’t flag anything though.

1 Like

Is it possible to make a note, and check the ‘ID not as good as it gets’ box?

Instead it’s the only way to deal with it fast, in school projects there’re thousands of observations every day and for some reason many iders don’t care about cultivated plants so they go RG fast, students also agree with each other’s ids, like in this case, asking for help doesn’t mean shaming organisation (though for sure it’s the fault of teachers, there’re easy ways to prevent this from happening, but they don’t care enough to do that or to learn about ways to do that).

2 Likes

I support asking identifiers for help. And it could be told to the staff, if the amount of misidentifications is interpreted to be large enough to help correct on our own. I’m mostly meaning not to flag anything. I do also think though that many IDs could get corrected eventually over time via new identifiers IDing them, vs. some seem to think misIDs will stay forever/be too numerous. So there’s the factor of revision over time. Of course, it helps to correct some of the misIDs asap to reduce the problem early.

So? Students are in school to learn; that means having their mistakes corrected. There is a reason schoolteachers have red pens. Whether or not the student likes their grade isn’t the point.

2 Likes

I apologise I was being a little vague here. Both cases we are battling against a project that has lots of problematic observations, but the differences between this one and the HK ones are that the latter just have loads of cultivated plants that are not marked casual, which is a far easier task compared to reverting tons of incorrect RG observations with 5 agreements. The Hymenoptera identifications are overall quite reliable, especially since we have some users very familiar with their morphology etc.

5 Likes

If we can’t be told what the project is, how can identifiers be recruited to do corrections?

3 Likes

I’d personally like to know what the project is - if it’s in my competences I’d be happy to help with correcting IDs in my free time

1 Like

Misids that get RG are likely to be left at this status for years if ever be corrected.

1 Like

Is it possible to filter within a project all observations that contain maverick IDs? I know this is possible for a specific user, but did not find a way to extend this search

1 Like

I’ve been keeping a very close eye on this thread from the start. It’s not perfectly in line with Forum rules in my opinion, but since the user is asking for methods – not necessarily direct help – I let it be.

We don’t allow links to misbehaviour (real or perceived) because the forum isn’t the place for that – flags are.
We don’t want people brigading or dogpiling.

Obviously if you were to try and brute-force the problem, you would need lots of help. But I doubt you can legitimately find eleven or fifteen people or however many who actually know the species.